2025Activity reportProject-TeamCOMBO
RNSR: 202524648M- Research center Inria Centre at Rennes University
- In partnership with:École normale supérieure de Rennes, Université de Rennes, Université Rennes 2
- Team name: Computer and Biomechanics –> Out-of-the-Lab
- In collaboration with:Institut de recherche en informatique et systèmes aléatoires (IRISA), Mouvement, Sport, Santé (M2S)
Creation of the Project-Team: 2025 January 01
Each year, Inria research teams publish an Activity Report presenting their work and results over the reporting period. These reports follow a common structure, with some optional sections depending on the specific team. They typically begin by outlining the overall objectives and research programme, including the main research themes, goals, and methodological approaches. They also describe the application domains targeted by the team, highlighting the scientific or societal contexts in which their work is situated.
The reports then present the highlights of the year, covering major scientific achievements, software developments, or teaching contributions. When relevant, they include sections on software, platforms, and open data, detailing the tools developed and how they are shared. A substantial part is dedicated to new results, where scientific contributions are described in detail, often with subsections specifying participants and associated keywords.
Finally, the Activity Report addresses funding, contracts, partnerships, and collaborations at various levels, from industrial agreements to international cooperations. It also covers dissemination and teaching activities, such as participation in scientific events, outreach, and supervision. The document concludes with a presentation of scientific production, including major publications and those produced during the year.
Keywords
Computer Science and Digital Science
- A5.1.4. Brain-computer interfaces, physiological computing
- A5.1.9. User and perceptual studies
- A5.6. Virtual reality, augmented reality
- A5.10.5. Robot interaction (with the environment, humans, other robots)
- A6.1.4. Multiscale modeling
- A6.5.1. Solid mechanics
- A9.2.1. Supervised learning
- A9.2.2. Unsupervised learning
- A9.2.4. Optimization and learning
- A9.3. Signal processing
- A9.12. Computer vision
- A9.12.1. Object recognition
- A9.12.4. 3D and spatio-temporal reconstruction
- A9.12.5. Object tracking and motion analysis
- A9.12.8. Motion capture
Other Research Topics and Application Domains
- B2.2.7. Virtual human twin
- B2.5.1. Sensorimotor disabilities
- B2.8. Sports, performance, motor skills
- B5.1. Factory of the future
- B5.6. Robotic systems
- B9.4. Sports
- B9.5.5. Mechanics
1 Team members, visitors, external collaborators
Research Scientists
- Nolwenn Fougeron [INRIA, Researcher]
- Franck Multon [INRIA, Senior Researcher, HDR]
Faculty Members
- Charles Pontonnier [Team leader, ENS RENNES, Associate Professor, HDR]
- Nicolas Bideau [UNIV RENNES II, Associate Professor]
- Benoit Bideau [UNIV RENNES II, Professor, HDR]
- Armel Cretual [UNIV RENNES II, Associate Professor, HDR]
- Georges Dumont [ENS RENNES, Professor, HDR]
- Diane Haering [UNIV RENNES II, Associate Professor]
- Pierre Hellier [Univ Rennes , Professor, HDR]
- Hugo Kerhervé [UNIV RENNES II, Associate Professor]
- Richard Kulpa [UNIV RENNES II, Associate Professor, HDR]
- Fabrice Lamarche [UNIV RENNES, Associate Professor]
- Caroline Martin [UNIV RENNES II, Associate Professor]
- Guillaume Nicolas [UNIV RENNES II, Associate Professor]
- Nicolas Vignais [UNIV RENNES II, Professor, HDR]
Post-Doctoral Fellows
- Annabelle Limballe [UNIV RENNES II, from Jul 2025]
- Aurelie Tomezzoli [INRIA, Post-Doctoral Fellow]
PhD Students
- Ahmed Abdourahman Mahamoud [INRIA]
- Louis Arles [UNIV RENNES II]
- Kilian Bertholon [UNIV RENNES II]
- Clelia Boulati [INRIA]
- Léo Denouel [UNIV RENNES II]
- Mael Gallois [INRIA]
- Leo-Paul Huar [INTERDIGITAL, CIFRE]
- Shubhendu Jena [INRIA, until May 2025]
- Elisa Jolas [Univ Lyon 1]
- Elyse Larribau [ENS RENNES, from Sep 2025]
- Guillaume Le Guludec [UNIV RENNES, from Jul 2025]
- Guillaume Loranchet [INTERDIGITAL, CIFRE]
- Lola Masson [UNIV RENNES II]
- Florian Matéos [UNIV RENNES II]
- Simon Ozan [UNIV RENNES II]
- Nolwenn Poquerusse [UNIV RENNES II]
- Suzon Pucheu [ENS RENNES]
- Valentin Ramel [INRIA]
- Nolwenn Regnault [UNIV RENNES II]
- Victor Restrat [INRIA]
- Etienne Ricard [INRS - VANDOEUVRE- LES- NANCY, until Sep 2025]
- Julie Robillard [UNIV RENNES II]
- Remy Roinson [Phyling,CIFRE]
- Sony Saint-Auret [INRIA]
- Aurelien Schuster [FONDATION ST CYR]
Technical Staff
- Guillaume Claude [UNIV RENNES II, Engineer]
- Ronan Gaugne [UNIV RENNES, Engineer]
- Laurent Guillo [CNRS, Engineer]
- Corentin Hardy [UNIV RENNES, Engineer, from Sep 2025]
- Julian Joseph [INRIA, Engineer]
- Antoine Marin [UNIV RENNES II, Engineer]
- Anthony Sorel [UNIV RENNES II, Engineer]
- Alexandre Vu [UNIV RENNES II, Engineer]
Interns and Apprentices
- Marcella Bergamini Wendhausen Portella [INRIA, Intern, from Jun 2025 until Aug 2025]
- Marie Curci [ENS RENNES, Intern, until Apr 2025]
- Erwan Fertray [UNIV RENNES, Apprentice]
- Elyse Larribau [CENTRALE NANTES, Intern, from Feb 2025 until Jul 2025]
- Minh Nhut Nguyen [UNIV RENNES II, Intern, from Mar 2025 until Aug 2025]
- Raphael Peus [UNIV RENNES II, Intern, from Feb 2025 until Jul 2025]
- Nicolas Polio [INRIA, Intern, from Feb 2025 until Jul 2025]
- Stephane Salvan [UBO, Intern, from Apr 2025 until Jun 2025]
Administrative Assistants
- Nathalie Denis [INRIA]
- Sophie Maupile [CNRS]
External Collaborator
- Mathieu Ménard [IO Rennes]
2 Overall objectives
2.1 Context and Motivations
Human beings are increasingly interacting with complex, connected and responsive systems. In sport, for example, athletes are increasingly using virtual reality systems to optimize their training 67. In the field of disability, exoskeletons and mixed reality systems are gradually being used to support rehabilitation or assist people with reduced mobility 68, while in ergonomics, these systems reduce the arduousness of work tasks while requiring complex motor adaptations 69. Humans may also have to carry out tasks in collaboration or in opposition with other humans, sometimes in hybrid environments mixing real and virtual humans, which is also a complex form of interaction from a biomechanical and sensorimotor point of view 66.
The behavior of these systems is increasingly well modeled, thanks in particular to the development of computing capacity. However, understanding and analysing the interaction between humans and systems requires detailed modelling of human behaviour in terms of motion and forces exchanged with the environment, in other words the biomechanical behavior. What's more, this understanding can only be achieved through field experiments to ensure that the analysis is ecologically valid. This need coincides with the arrival of mature capabilities for measuring movement in the field (IMUs, video capture, markerless, etc.). These tools open the door to systematic monitoring of sporting or professional activity, or monitoring of the development of a disability, among many other applicative examples.
The maturity of measurement methods in the field, combined with the increasing complexity of human-system interactions as a result of technological development, has led to a growing scientific need for analysis, classification and simulation of human movement and interaction from a biomechanical point of view.
2.2 Scientific objectives
Based on this observation, the ComBO project team will aim to develop biomechanical models and numerical methods for analyzing and simulating human-system interaction on field. We define ourselves as digital biomechanicians, which is a unique skill within Inria, while having strong connections with the fields of robotics and virtual reality. To illustrate the objectives of the team, we can draw on concrete applications, as shown in Figure 1:
- Example 1 We want to understand how an athlete interacts with his sports equipment, in this case how he exploits the pole vault in terms of energy storage and restitution, which is a criterion for improving fundamental performance in pole vaulting. This first example raises a number of scientific challenges. First of all, there is the need for a realistic, detailed and personalised biomechanical model of the athlete in action, enabling detailed information to be retrieved about the athlete's joint and muscle mobilisation during movement. This model also needs to interact with a model of the sports equipment in a simulation that can be controlled using movement synthesis and analysis. In this particular case, the pole is a passive, highly flexible system that presents inherent difficulties to be properly modeled and simulated. Finally, to feed this simulation and then validate it, it is necessary to capture and analyse the movement of the pole vaulter and the pole vault in a real situation, which poses clear limits on the quantity and quality of the signals captured and the necessity to learn from a more complete data a sparse response (see Figure 1.a).
- Example 2 We want to assess and train boxers' dodging skills while limiting the physical impact of this training, and train against a given style associated to a future opponent. To this end, it is necessary to extract, model and simulate the fighting style, including motion and strategy, observed in real condition. It raises several scientific challenges, such as collecting 3D data of real fights, including shape and motion of the two interacting fighters, with limited equipment. A second challenge is to define a relevant, compact and controllable representation of the fighting interaction, with temporal and spatial dependencies, for unstructured clips. This model enables to simulate a virtual opponent in a mixed reality system, but this system should lead to realistic behaviours of the user when fighting against this model, to ensure transfer of skills between virtual training to real fighting (see Figure 1.b).
- Example 3 We want to be able to investigate and quantify the biomechanical impact of an assistance system on the performance of work tasks. In this example we find the need for a detailed musculoskeletal model, in this case of a specific joint, and also the challenge of co-simulating the human and the assistance system and their respective control (feedforward and feedback strategies). An additional scientific difficulty lies in the need to model the contact actions at the interface between the human and the exoskeleton, which are key to any simulation seeking to assess the advantages and disadvantages of human interaction with such systems in biomechanical terms (see Figure 1.c).
- Example 4 We want to identify the actual motor strategy of disabled people, in particular in daily tasks involving upper body. The challenge is to distinguish between capacity, i.e. physiological possibility to perform such gesture, and actual usage in daily life. During clinical evaluation, disabled people will tend to act as they think health professional want them to act whereas the main question is to know how they act in real life (see Figure 1.d).
Four illustrations of the team work. The first illustration is a pault volter equipped with sensors, the second a boxer equipped with a VR helmet, the third a wrist exoskeleton equipped on a butcher, the last a child in front of a capacity benchmarking protocol, assessing its motor capabilitées.
Four illustrations of the team work. The first illustration is a pault volter equipped with sensors, the second a boxer equipped with a VR helmet, the third a wrist exoskeleton equipped on a butcher, the last a child in front of a capacity benchmarking protocol, assessing its motor capabilitées.
3 Research program
The team is developing a three-tiered methodology, enabling it to develop different answers to the questions raised by the applications. In particular, we are considering three different working environments that will enable us to develop our scientific approaches (see Figure 2):
- The laboratory This is the natural place for prototyping methods, where we will find the most powerful technologies for capturing effort, movement, muscular activity, etc. It is also the ideal place to carry out a analyses of gestures under controlled conditions.
- The XR Extended reality (VR and AR) is an ideal tool for analysing movement, where control can be high while offering sensory-motor alteration strategies that may be useful for training or prevention. It's also a "safe" place to evaluate certain practices in secure conditions.
- The real world This is where the previously developed methods are applied and analyses are carried out with the application in mind. This is also generally where the motivation for the scientific work that lead to the development of the original methods developed by the team is to be found.
These three environments present an increasing level of ecological validity while they present a decreasing level of control (in terms of experimental conditions and motion and force sensing). One of the challenges is therefore to progressively reduce the level of control over the developments carried out, in order to get closer to the conditions of field applications. The cross-sectional approach of the team is leading to four research axes that are summarized here and developed in the next sections of the document:
- Ready-to-go Biomechanical Models: The current maturity of musculoskeletal modelling and personalizing tools, and the democratisation of medical imaging data, mean that in the near future it will be possible to create detailed biomechanical models of humans that can be personalised with just a few measurements and a few clicks. This objective aim at representing specific subjects (high-level athletes, specific disabilities), or particular populations (workers in a given sector, age cohorts, etc.). To achieve this objective, it is necessary to develop both statistical learning methods to reduce the specific data required to establish a model representative of a given population, and model-informed methods based on anatomical and physiological knowledge to obtain unique characteristics representative of the subject's physical capabilities.
- Efficient Extended Environments: The emergence of immersion tools such as Head-Mounted Displays (HMDs) and haptic interfaces makes it possible to envisage the creation of virtual or augmented environments that can be used to study the physical activity of interacting humans in an efficient manner. In particular, sports applications of this type raise a number of scientific questions, including the quality of immersion, the fidelity of physical interactions with the environment, and the evaluation of training methods using this paradigm (skills training), in particular regarding the biases integrated into the perception-action loop (perception of one's body, shift in sensory stimuli, etc.). Such questions may also be developed for other applications (ergonomics, clinics).
- Human-System Interaction: The development of computing capacity and the associated algorithms make it possible to implement detailed co-simulations of humans and the systems with which they interact, whether from a sports (athlete-sports equipment), ergonomics (worker-assistive device) or clinical (patient-medical device) point of view. In particular, these simulations and the study of the resulting interactions can be used to predict the impact of systems on the performance of a given task, thereby accelerating their prototyping or optimizing their use. These objectives raise a number of scientific challenges, such as modelling the human-system interface, integrating control into the simulation and rigid-deformable coupling in the simulation loop. It also raises questions about the interpretability of the results and their analyses.
- Ready-to-go Field Analysis: Field analysis methods will become more widespread in the future, particularly for monitoring athletes during their training. In the long term, these analysis methods could lead to significant improvements in training, whether in terms of physical preparation, injury prevention or optimisation of the sporting gesture. This perspective can be transposed to physical activity in the workplace for injury prevention, or for monitoring patients on a daily basis. However, the accuracy and quantity of the data available mean that this analysis is very partial at present, and requires the development of methods based on learning and biomechanics to increase the data available and maximise the relevance of the approach for the practitioner. The latter also requires the tools to be designed from a field perspective, by co-constructing the analysis tools with those practitioners.
The picture shows 3 levels of work, one being the lab (illustrated by a subject in a lab configuration for motion capture), the second being the XR (a boxer with a VR helmet), the third being the field (a swimmer in action).
4 Application domains
4.1 Applicative fields
The team contributes to numerous numerical topics due to its multidisciplinary nature: human movement analysis, computer vision, biomechanical system simulation, virtual reality, biofeedback, motion recognition.
We are also involved in three cross-disciplinary fields, which can be summarized as follows:
- Sport: Sport is one of the major application field of the team. Studying the biomechanical constraints suffered by the athlete during his/her performance enables to identify the possible injury mechanisms associated to a gesture and to adapt the training and the physical conditioning in relation with those mechanisms. This observation is directly translatable to the performance itself, the human-material interaction being directly optimized through this biomechanical analysis. Such analyses may be done in an acute manner, but can as well be considered on a longer duration, by evaluating and analysing this interaction during the whole athlete preparation with a proper field analysis protocol. Furthermore, the use of new training medias such as extended reality express the need to understand the biomechanical behavior of the athlete in this environment, and to quantify the interaction with the system in relation with his/her sensorimotor behavior. Last, disabled sports need a heavy investment in the understanding and the optimization of the athlete-material interaction which is still largely suboptimal due to the relative confidentiality of this activity and the lack of investment in a more general way.
- Ergonomics: Ergonomics, and in particular the analysis of human physical activity at work, is another of the team's key areas of application. The study of the biomechanical constraints associated with the performance of professional tasks remains a key factor in occupational health, particularly in the prevention of musculoskeletal disorders. In addition, the complex interaction between humans and the machines that surround them in a professional context, for example in industry, requires special treatment to understand the movements and efforts involved in this interaction. This analysis is necessarily carried out under the constraints of the field, which will reduce the quality of the data that can be captured and necessitate the development of methods for estimating specific biomechanical quantities. Finally, the advent of work assistance systems (cobots, exoskeletons) requires a detailed understanding of human-system interaction in a professional context, in order to design high-performance systems and assess their contribution at the end of the chain.
- Clinics: The team also deals with scientific problems that have their origins in clinical practice. In particular, the team is interested in analyzing the physical activity of people with disabilities. There is still a major need to develop field tools for quantifying biomechanical parameters as part of the treatment of a pathology, in order to help therapists make the right decisions. This need goes hand in hand with the need to develop models specific to the pathology studied (neurodegenerative diseases, post-avc, etc.). In addition, the number of rehabilitation and assistance systems is increasing, and their effectiveness has yet to be evaluated in the short and long term. The team is therefore also seeking to assess the biomechanical constraints associated with the use of such devices in a therapeutic context. The problems of the field are again present here, since the effectiveness of such systems can only really be assessed in everyday situations.
4.2 Scientific impact
The team will contribute to extend the scientific knowledge on several aspects:
- All of our developments in biomechanics, especially in human motion analysis, synthesis and modeling are dedicated to better understand human motion and human interaction in our more and more complex environments. Through our developments of data-driven (learning based) approaches all along the motion analysis/synthesis pipeline, we contribute to the development of a new level of description for human motion, enabling a better understanding of the mechanisms underlying human activity from a biomechanical point of view, in its environment and interacting with it;
- In numerical fields, the extension of existing learning-based (machine learning, deep learning) methods to experimental, physiological data is a strong scientific asset, in which we will be able to contribute heavily to generalize these approaches to sparse, small sample sizes and noisy data;
- To go beyond the development of human digital twins, we aim at developing innovative coupled models taking humans and their environment into account. Indeed, the development of novel human-system simulation frameworks is also an important stake, since all the main actors in biomechanics, sport science and robotics seek at representing the human in the simulation loop to enhance design, control, use and performance of the systems, being assistive devices, sport material or virtual reality headsets for example. The team, with its strong expertise in biomechanics and the development of its expertise in coupling human body models and systems, aim at having a major role in the development and democratization of such tools;
- Our developments around the use of extended environments (VR/AR/XR) for physical activity assessment is also contributing to the understanding of our behavior with regard to these technologies, and reversely to adapt these technologies to human being. Such an understanding is of primary importance with regard to the place it will take in the following years, with the dissemination of "Metaverse" applications, such as educational, professional and sports training, rehabilitation, physical activity assessments...
- In sports sciences, all the developments of the team contribute to better understand performance, injury, or recovery mechanisms with regard to the interaction of the athlete with his/her material or his/her environment. This is the main application of the methods developed by the team and should help, applied to specific disciplines, to extend the knowledge related to a given sport gesture or the relation between sport performance and physical condition or sport performance and human-material interaction for example;
- In studying physical activity at work (physical ergonomics), our work will similarly enhance the knowledge of the biomechanical constraints related to the work gestures in specific conditions. Our focus on physical assistance will also have a strong impact on the understanding of the interaction between human and systems, and ways to maximize the positive effects of such an assistance from a design, control and use point of view. It will contribute to better take human factors in the next generations of manufactures centered on working conditions;
- Finally, in clinical the development of specific models of disabled people will be a first asset of importance, to better understand physical constraints related to a given pathology and how to deal with it from a clinical and patient point of view. As well, focusing on assistive devices also seek at enhancing the knowledge around the effects of such systems at the kinematic, dynamic and muscle level to maximize their efficiency.
5 Social and environmental responsibility
5.1 Societal and Economical Impact
The focus of the team on analyzing human activity on the field should be able to provide the best level of knowledge to the practitioner with a minimal intervention on the athlete/patient/worker performance. Proposing such analyses will have a very strong impact on the development of human motion quantification. This is a major societal stake since well being and health became a central element in the modern societies. The team already has a solid set of partnerships with industry, clinicians and sports federations in which we largely apply our expertise to concrete and useful solutions. We already contributed to assist sports training staff to prepare elite athletes for the Olympics. This activity will be continued to produce new knowledge and methods to enhance performance and preserve health for the next competitions. In the same way, we have created a start-up company named Moovency in 2017 to support ergonomists when analyzing the workers' activity on site. The ComBO team aims at continuing this very practical approach by continuing the transfer of its development to the society.
By continuing these partnerships and extending the applications, the team has a role to play in the democratization and the development of systems in interaction with humans. For example, our objectives in terms of human-system simulation should serve as an accelerator to the prototyping of assistive devices and their use in industrial or rehabilitation contexts. In a similar manner, our developments around efficient extended environments should play a major role in the development of XR solutions dedicated to i) learning skills including complex and distant collaborative tasks, ii) practicing physical activity in a standardized and safe environment, iii) rehabilitation purposes with a human-centric development scheme to design assistive devices and rehabilitation protocols. These elements are clearly related to training, health and well being issues, and should transfer to industrial partners to touch the largest population. Therefore the economical impact of the team will be linked to its capacity to transfer applications to industry, as it has already been the case in the past (see industrial partnerships).
Finally, at the era where everything can be measured through connected devices, developing feedback (real time or to practitioners) should also be an accelerator for the reliability and democratization of these systems to help any of us to enhance health practices. ComBO will have a unique collection of multidisciplinary skills to enhance these systems, make them available to the wide public audience, and propose smart interfaces to analyze and interpret these measurements.
5.2 Environmental Impact
The environmental impact policy of the team can be considered from two points of view:
- The team's travel policy should be exemplary. Beyond the rules common to Inria teams and the IRISA laboratory, air travel should be reserved for destinations that are truly essential to the life of the team (presentations at A* rank conferences, and exchange of experts) and designed to maximize the value of the trip (extension of the mission by laboratory visits, collaboration meetings, etc.). Those questions may have an impact on the collaborations the team can run, favouring "local" (Europe) partners instead of worldwide partners. We are aware that those element may have a negative impact on the research, but factually most of our collaborations are already in Europe.
- The team's equipment policy should be designed to encourage the use of common tools for experimental aspects, in vivo or in silico. In view of the growing need for computing time associated with the expansion of learning methods, the team will be keen to implement these calculations methodically, and after ensuring that the requested calculation is scientifically mature. The team will also need a number of in vivo measurement tools (optoelectronic motion capture, markerless, IMU, EMGs, effort sensors, etc.) and will draw on existing tools in the previous team (MimeTIC), the Inria center, IRISA and M2S to build a coherent usage policy. In particular, we will seek to propose a system for reserving experimental resources that will enable us to rationalize their use as much as possible. From a purchasing point of view, the team will seek to limit its environmental impact, in particular by avoiding hasty spending at the end of the projects, which is highly deleterious from this point of view. By supporting the immerStar platform, we also support any possibility to mutualise tools, software and equipment with other teams and laboratories, to prevent from buying them several times on the same site. This intention may have consequences on the research and the visibility of the research. We do not renounce at buying additional material, but this should be done with the idea of maximizing their damping both economically and environmentally. The main concern is that we should still be on page on the most recent technologies of motion capture. At this date, we are still using one of the most performing system (Qualisys), and we regularly bought additional cameras to enhance the platform. In future updates, we should consider the gain (that may sometimes marginal) with regard to the environmental cost.
5.3 Ethics
The team's overall aim is to develop ethical applications in areas such as occupational health, sports for all and disability assistance. This applicative vision goes hand in hand with ethical scientific behavior, which must be beyond reproach. The MimeTIC team, prior to the ComBO team, has long implemented an ethical policy associated with its experimental nature, by systematically submitting its experimental protocols to the appropriate committees, whether Inria's COERLE for applications involving hardware or methodological development, or the Comités de Protection des Personnes for applications requiring it (research on humans). This policy will be pursued and expanded, in particular by proposing joint protocols with players in the healthcare, sports and labor sectors. Generic referrals will be proposed to COERLE to facilitate the implementation of experiments and their ethical validation. Particular attention will also be paid to the collection and management of personal data, in line with GDPR requirements.
6 Highlights of the year
The projects REVEA and sharespace were presented at the AI For Good Global Summit 2025 in Geneva, supported by United Nations. They presented demos of :
- cycling scenario to anticipate an attack in a virtual peloton
- boxing against an AI opponent trained to imitate the style of a mocaped boxer
7 Latest software developments, platforms, open data
7.1 Latest software developments
7.1.1 CusToM
-
Name:
Customizable Toolbox for Musculoskeletal simulation
-
Keywords:
Biomechanics, Dynamic Analysis, Kinematics, Simulation, Mechanical multi-body systems, Musculoskeletal Analysis
-
Scientific Description:
The present toolbox aims at performing a motion analysis thanks to an inverse dynamics method.
Before performing motion analysis steps, a musculoskeletal model is generated. Its consists of, first, generating the desire anthropometric model thanks to models libraries. The generated model is then kinematical calibrated by using data of a motion capture. The inverse kinematics step, the inverse dynamics step and the muscle forces estimation step are then successively performed from motion capture and external forces data. Two folders and one script are available on the toolbox root. The Main script collects all the different functions of the motion analysis pipeline. The Functions folder contains all functions used in the toolbox. It is necessary to add this folder and all the subfolders to the Matlab path. The Problems folder is used to contain the different study. The user has to create one subfolder for each new study. Once a new musculoskeletal model is used, a new study is necessary. Different files will be automaticaly generated and saved in this folder. All files located on its root are related to the model and are valuable whatever the motion considered. A new folder will be added for each new motion capture. All files located on a folder are only related to this considered motion.
-
Functional Description:
Inverse kinematics Inverse dynamics Muscle forces estimation External forces prediction
-
Publications:
hal-02268958, hal-02088913, hal-02109407, hal-01904443, hal-02142288, hal-01988715, hal-01710990, hal-03279707v1
-
Contact:
Charles Pontonnier
-
Participants:
Suzon Pucheu, Antoine Marin, Diane Haering, Aurelie Tomezzoli, Aurelien Schuster, Antoine Muller, Charles Pontonnier, Georges Dumont, Pierre Puchaud, Anthony Sorel, Claire Livet, Louise Demestre
7.2 New platforms
Immerstar
Participants: Ronan Gaugne, Georges Dumont, Antoine Marin, Anthony Sorel, Richard Kulpa.
With the two platforms of virtual reality, Immersia and Immermove Immermove, grouped under the name Immerstar, the team has access to high level scientific facilities. This equipment benefits the research teams of the center and has allowed them to extend their local, national and international collaborations. The Immerstar platform was granted by a Inria CPER funding for 2015-2019 that enabled important evolutions of the equipment. In 2016, the first technical evolutions have been decided and, in 2017, these evolutions have been implemented. On one side, for Immermove, the addition of a third face to the immersive space, and the extension of the Vicon tracking system have been realized and continued this year with 23 new cameras. And, on the second side, for Immersia, the installation of WQXGA laser projectors with augmented global resolution, of a new tracking system with higher frequency and of new computers for simulation and image generation in 2017. In 2018, a Scale One haptic device has been installed. It allows, as in the CPER proposal, one or two handed haptic feedback in the full space covered by Immersia and possibility of carrying the user. Based on these supports, in 2020, we participated to a PIA3-Equipex+ proposal. This proposal CONTINUUM involves 22 partners, has been succesfully evaluated and will be granted. The CONTINUUM project will create a collaborative research infrastructure of 30 platforms located throughout France, to advance interdisciplinary research based on interaction between computer science and the human and social sciences. Thanks to CONTINUUM, 37 research teams will develop cutting-edge research programs focusing on visualization, immersion, interaction and collaboration, as well as on human perception, cognition and behaviour in virtual/augmented reality, with potential impact on societal issues. CONTINUUM enables a paradigm shift in the way we perceive, interact, and collaborate with complex digital data and digital worlds by putting humans at the center of the data processing workflows. The project will empower scientists, engineers and industry users with a highly interconnected network of high-performance visualization and immersive platforms to observe, manipulate, understand and share digital data, real-time multi-scale simulations, and virtual or augmented experiences. All platforms will feature facilities for remote collaboration with other platforms, as well as mobile equipment that can be lent to users to facilitate onboarding. The kick-off meeting of continuum has been held in 2022, January the 14th. A global meeting was held in 2022, July the 5th and 6th.
7.3 Open data
8 New results
8.1 Axis 1: Ready-to-go biomechanical models
8.1.1 Subject specific modeling
Those results deal with subject specific modeling at the osteoarticular and muscular level.
Subject specific lower limbs and knee modeling
Participants: Nolwenn Fougeron, Charles Pontonnier, Elisa Jolas.
Knee representation including manual and automatic landmarks.
In preoperative planning, a number of techniques for registration of anatomical landmarks on bones, or “bones landmarking”, have been developed for automatic measurement of lower limbs. However, the evaluation of these techniques is often neglected, reducing confidence in results and limiting clinical applications. We propose a protocol to evaluate landmarking techniques applied to thirty 3D images of osteoarthritic lower limbs. The protocol includes the evaluation of the ground truth based on manual landmarking performed by two experienced clinicians. Intra- and inter-operator variability were quantified. Automatic landmarking was evaluated using the same procedure and compared to manual landmarking by computing distances with bounding boxes including manual landmarks. The method was applied to 87 landmarks on the femur, tibia, fibula, and patella. For intra-operator variability, the mean median error was 2.0 mm, with a maximum of 42.1 mm. For inter-operator variability, errors ranged from 0.0 to 43.5 mm, with a mean median error of 2.3 mm across all landmarks. The mean median error between manual and automatic landmarking methods was 2.4 mm, with a range from 0.0 mm to 22.8 mm. Confidence intervals of automatic landmarking errors were consistent with intra- and inter-operator variability confidence intervals. Approximately 42 % of the automatic landmarks were inside their bounding boxes, with a mean distance of 1.4 mm from the box for landmarks on the outside. Results showed different errors depending on landmark location and coordinates. Intra- and inter-operator variability shed light on the difficulty of defining ground truth based on manual segmentation. This finding emphasizes the need for a rigorous evaluation protocol for landmarking validation, including the verification of statistical test hypotheses and population size influence. Published in 15.
In addition to this major contribution, many works in the team focuses in improving knee modeling, especially in adding complexity in the musculoskeletal models (ligaments, moving rotation axes)...It has been published in different short communication in congresses for now 62, 12, 48, 16, 50, waiting for extended results to be published in journal papers.
Subject specific force generation capacities modeling
Participants: Charles Pontonnier, Aurelie Tomezzoli.
Torque-angle curves including clinical scores
Within the frame of the MusMapS exploratory action, we seek at defining subject-specific force generation capacities models for people with motor impairment (stroke, multiple sclerosis). In collaboration with the Fondation Saint Hélier (rehabilitation center in Rennes), we evaluated the strength developed by such a population in shoulder internal/external rotation and elbow flexion using an isokinetic ergometer (a rehabilitation machine that provides torques developed by the patient during a given exercise). Using those data, we are able to i) provide reliable torque-angle relationships for these people 57 and also to compare and evaluate this data to the clinical scores provided by practitioners in their evaluation of the patients 56, 30. Such a work will be of first importance for the next steps of the project, in which we will exploit this data to provide personnalized as-needed assistance to users with an upper limb exoskeleton.
8.1.2 Motion analysis
These results deal with human motion analysis in a laboratory setup, including applications in sports, ergonomics and clinics.
Stepping analysis
Participants: Charles Pontonnier.
No step strategies are differentiated from backwatd and forward steps. Lateral steps are classified in loaded side step, unloaded medial step and unloaded crossover step.
Recovery from external perturbations typically involves stepping, with the perturbation direction playing a key role in determining the recovery strategy. To date, classifications of these stepping strategies have relied on prior knowledge of perturbation direction, which is not always available when considering experimental paradigms close to real-world scenario. Here, we introduce a novel Unified classification method that enables the labeling of first recovery steps based solely on body kinematics. We have also developed and validated a logistic regression model that effectively differentiates between different recovery strategies. Published in 8, 61. A collaboration with the VirTUS Inria research team.
Motion analysis and chronic low back pain
Participants: Mathieu Ménard, Diane Haering, Armel Cretual.
Chronic non-specific low back pain (cNSLBP) is a leading cause of disability, influenced by bio-psycho-social factors. However, its impact on everyday activities such as navigating streets and interacting with other pedestrians remains underexplored. This study aimed to assess the effect of cNSLBP on perceptual-motor processes in a pedestrian crossing task, focusing on 1) collision avoidance behaviours, 2) the walker's role in avoiding collisions, and 3) the influence of pain perception. Methods: Seventeen asymptomatic adults (AA, 11 females, 46.4 ± 12.8 years) and seventeen cNSLBP participants (10 females, 47.9 ± 12.7 years) performed a task involving crossing paths at a 90° angle with another walker. Participants interacted in three groups pairings: AA-AA, AA-cNSLBP, and cNSLBP-cNSLBP. Key metrics included crossing order inversion, collision risk threshold informing movement adaptation, crossing distance, and the walker's contribution (speed/orientation). Results and discussion: No significant differences were observed between groups for the collision risk threshold (0.93 m) or crossing distance (0.8 m). However, cNSLBP participants exhibited distinct avoidance strategies, especially in cNSLBP-cNSLBP interactions, which showed more frequent inversions. When crossing first, cNSLBP participants contributed less, whereas when crossing second, they contributed more, primarily by adjusting their speed. A significant negative correlation emerged between depression scores and the level of contribution when cNSLBP participants crossed second. Conclusion: These findings suggest that pain perception may influence collision avoidance behaviours. Further research, potentially incorporating virtual reality, is needed to control environmental factors and deepen our understanding of these interactions. Published in 7, 60, in collaboration with the VirTUS Inria research team.
Additional research is led on chronic low back pain in the team, mainly focusing on the efficiency of chiropractic therapies on the low back pain (see 11).
Motion analysis and sport sciences
Participants: Nicolas Vignais, Anthony Sorel, Nolwenn Poquerusse.
We had various results in the field of motion analysis for sport sciences. For example, we developed additional results on discriminating sensorimotor performance in elite football players by evaluating their reactive and planned agility 22 (a collaboration with the Stade Rennais football team). We also had some studies about studying cycling motion with IMU sensors , showing that they are reliable to asses joint torques against time 46, 45.
Work in progress and books contributions
Participants: Nicolas Vignais, Aurelie Tomezzoli, Charles Pontonnier, Georges Dumont, Aurelien Schuster, Franck Multon.
We also had some interesting collaborations leading to the evaluation of the impact of the bow camber on violonist kinematics 37 (with Université Lyon 1). We also evaluated the possibility to use mecanomyography to assess muscle fatigue, developing specifing signal artifacts removing methods 35, 47. We participated to the publication of a reference book for sport sciences students, presenting biomechanics and human motion analysis for undergruate students 65. Lastly, we proposed a mechanical energy analysis of obstacle crossing in soldiers, enabling a better understanding of this crossing technique and its biomechanical implications 55.
8.2 Axis 2: Efficient extended environments
8.2.1 Studying perception-action loop in virtual reality for sports applications:
The perception-action loop has long been studied in sports, to explain athletes decisions and actions with regard to their direct environment and the clues they perceive from the opponents actions. The team continues to develop works in this field.
Visual Tracking Performance of Soccer Players in a Virtual Reality Multiple-Player-Tracking Task
Participants: Alexandre Vu, Anthony Sorel, Benoit Bideau, Richard Kulpa.
A participant in front of a cave emulating a virtual soccer field.
Visual tracking tasks are commonly used in laboratory settings to investigate the relationship between team sports expertise and the ability to dynamically allocate attention to multiple moving targets. It has been studied for football in the present study 64, especially the interaction of this tracking performance with the reception of a pass.
Custom Virtual Reality Exergame for Hospitalized Children with Cancer
Participants: Alexandre Vu, Benoit Bideau.
a. Start gamescreen in sky environment; b. Game session with obstacles; c. Child playing the VR-based exergame.
Exercise improves health outcomes in hospitalized children with cancer but is often limited by various barriers. Virtual Reality (VR) based exergames offer a promising solution. This study evaluated the acceptance, compliance, and intensity of a custom VR-based exergame program for pediatric cancer patients in a hospital setting. Thirteen children participated, achieving a compliance rate of 68% and an average intensity of 4.92 Metabolic of tasks (MET), meeting recommended moderate-to-vigorous physical activity (PA) levels. The findings suggest that this VR-based exergame is a promising and engaging intervention to promote exercise in children with cancer during treatment, with minimal risk of cybersickness. Published in 36.
Physical bounce model for virtual racket interactions
Participants: Sony Saint-Auret, Franck Multon, Ronan Gaugne, Richard Kulpa.
Experimental set-up composed of a racket rigidly attached to a support, the ball with coloured patches (detailed Jeu de Paume balls on the right) going out of a tube held at different heights, orientations and positions.
Nowadays, Virtual Reality is widely used in sports, to enhance physical fitness, or improve specific subskills, such as anticipation skills. However, many factors in VR can alter the experience and make it difficult to transfer the skills trained in VR to real practice. One of these factors is the physical simulation of the virtual environment, that may produce unexpected behaviours. Hence, if users are athletes in ball-based sports, the VR training simulator should compute ball trajectories that look plausible for them. In this paper, our aim is to evaluate how human perception can be influenced by variations in a ball physics' model. We explore properties of human perception, the acceptance threshold beyond which a deviation from the reference ball trajectory is perceived more than 50% of time, and the Just-Noticeable Difference (JND) as an indicator of perceptual sensitivity. To this end, we conducted psychophysical experiments where participants were asked to either only observe, or observe and hit virtual bouncing balls simulated with varying coefficients of restitution (COR). We report the acceptance threshold and JND in different conditions. We found that participants detected variations in COR more easily when having the motor task. Additionally, their sensitivity to variations was globally higher when they first performed the perceptual task alone, before the motor task was introduced. These results contribute to the design of credible VR environments involving bouncing objects, such as for virtual sports. Published in 43.
8.2.2 Virtual reality for arts and history
These results deal with the use of virtual environments to assess human behavior in relation to arts or historical questions, that are partly tackled within the team.
Reconstructing Gladiator Combat in A Multisensory Virtual Reality Training Environment
Participants: Stephane Salvan, Charles Pontonnier, Ronan Gaugne.
A hall with a 3D representation of a gladiator (provocator) and its equipment.
This study presents a research project focused on designing, implementing, and evaluating a multisensory virtual environment to simulate gladiatorial training. The aim is to analyze how immersive experiences impact the acquisition and refinement of technical skills in armed singular dueling. Conducted collaboratively by teams in virtual reality, biomechanics, and history, the project developed a historically contextualized environment centered on the provocator, a specific gladiator type. The virtual environment allows users to train in typical offensive maneuvers, offering a testbed for hypotheses about Roman combat and the effects of external conditions on performance. It serves as both a historical reconstruction tool and an experimental platform for studying ancient martial techniques. Built on rigorous historical and visual research, it uses motion capture technology to accurately recreate combat sequences, enhancing the authenticity and educational value of the simulation. A key contribution of this work lies in advancing the study of gladiatorial techniques, an area often distorted by popular culture. By integrating passive haptic and auditory feedback, the environment enhances sensory immersion, contributing to a deeper and more accurate understanding of gladiatorial practices. This multisensory approach not only supports the preservation of ancient techniques but also sheds light on the physical and cognitive demands faced by historical fighters. Ultimately, this research bridges disciplines-combining historical scholarship, biomechanics, and virtual reality-to offer an innovative way of exploring Roman gladiatorial training. The findings may inform broader discussions on the role of immersive technologies in skill development and historical interpretation within virtual environments. Presented and published in 40, 34 and as part of the developments currently made in the Immersia virtual reality cave 39.
Le Voyage du Geste
Participants: Ronan Gaugne.
'Le Voyage du Geste' project is part of an immersive and interactive artistic creation process in virtual reality, at the crossroads of digital art, dance and sensory experience. It aims to design a collaborative virtual reality (VR) work in which body movement – hand gestures – becomes both a medium of expression and a tool for transforming the virtual landscape. Inspired by the aesthetics of pictorial romanticism, this work invites viewers on a contemplative journey in which the participants' gestures shape mountains, flows of light and sound variations in a sensitive and poetic choreographic dynamic.
Conceived in the context of supportive care for women with metastatic breast cancer, Voyage du Geste seeks to demonstrate that a deep artistic commitment can become a lever for therapeutic support. The work invites patients to reconnect with their bodies through guided or free and intuitive dances, where each gesture produces sensitive effects in the virtual landscape. This gentle, progressive and adapted artistic approach to movement aims to reduce kinesiophobia while promoting relaxation, the pleasure of movement and letting go. By reintegrating movement into an aesthetic and immersive experience, the project offers a new form of ‘Poetic’ physical activity that is personal, emotional and potentially beneficial to quality of life. Presented in 63.
8.2.3 Work in progress:
Participants: Clelia Boulati, Julie Robillard, Anthony Sorel, Nicolas Vignais.
In a preliminary study, we realized an analysis of spatiotemporal parameters involved in the Lineout in rugby union 44. This preliminary analysis was necessary to consider modeling a virtual environment to develop the thrower performance based on subtle changes in the lineout organisation against time.
A work in progress of the PhD thesis of Julie Robillard has been presented in a conference 53. Through this doctoral project, a virtual environment will be modelled based on an industrial work situation, and the behaviour of operators in this environment will be recorded and then compared with that obtained in the real work situation. More specifically, a set of inertial measurement units (MVN Awinda, XSens, Movella) and instrumented gloves will be used to obtain the operators' kinematic data during the real work situation. The virtual environment will then be designed based on the workstation studied. The aim of this comparison is to position the use of virtual reality as a benchmark method for ergonomic analyses carried out prior to the design of workstations. A sensitivity study will therefore be conducted to identify the immersion parameters that have the greatest influence on operator behaviour in virtual environments. Based on the results obtained during the two previous stages, the virtual environment will be optimised to study operators' familiarity with this type of innovative methodology.
8.3 Axis 3: Human-system interaction
8.3.1 Interaction forces analysis and prediction
In human-system biomechanical,interaction, it is of primary importance to measure or estimate the external forces applied to the subject you are currently studying, since it enables the evaluation of internal forces. The ComBO team has a long experience of force prediction methodologies to predict external forces in absence of force sensor.
Size-adjustable instrumented track for reaction forces measure during running
Participants: Nicolas Vignais, Remy Roinson.
a 3D scheme of the instrumented track detailing sensors.
Force plates are widely used in sport sciences to measure ground reaction forces. However, their use in ecological conditions is limited by their restricted recording measurement area. Recently, a size-adjustable track instrumented with force sensors has been developed to measure vertical ground reaction forces (vGRF) over large perimeters. The aim of this study was to evaluate accuracy and repeatability of this instrumented track (IT). The evaluation was conducted through a two-step protocol: (i) a static protocol assessing the accuracy and repeatability of force sensors composing the IT, and (ii) an ecological protocol aiming to assess values obtained while running on both a reference force plate and a three-sensor IT configuration. Results from the static protocol demonstrated an excellent agreement between the reference force sensor and the IT sensors (p < 0.001, ICC(3,1) > 0.99). Force measured from the IT sensors showed a high accuracy with an average bias of 1.27 N, corresponding to a ratio limit of agreement of 0.3%, and a repeatability error within 2 N. For the ecological protocol, there was also an excellent correlation between the vGRF running parameters measured by the reference force plate, and the IT (p < 0.001, ICC(3,1) > 0.98), reinforced high accuracy results. Moreover, a ‘Minimal Detectable Changes’ analysis indicated that IT can accurately quantify all running patterns. This study demonstrated that force sensors integrated into the IT may be used to assess running patterns over a large area, expanding the scope of dynamic analysis in various applications. Published in 29. A side result about center of mass predicition has been presented in 28. A tentative work about predicting antero-posterior and medio-lateral forces from the vertical component during running has been presented in 54.
Acceleration prediction and analysis in gymnastic landing
Participants: Suzon Pucheu, Diane Haering, Mathieu Ménard, Charles Pontonnier.
Gymnastic landing is well known source of injuries in this sport, leading to serious sprains at the ankle, or anterior cruciate ligament ruptures, or low back pains. The stresses experienced by athletes during landings are a major issue in injury prevention and improving landing techniques. We have been working on these issues for several years.
In our ongoing work on this subject, we have developed a method for learning about the acceleration experienced using an LSTM-type neural network, based on a database collected from an instrumented impactor and measurements of the accelerations of the impactor and the mat. Published in 23.
We have also carried out an analysis of the impacts experienced by athletes during standardised gymnastics training, showing a significant increase in the accelerations experienced during training, which can be interpreted as a decrease in the shock absorption capacity of gymnasts during landings due to peripheral fatigue. Published in 9.
Ground Reaction Forces estimation method for non-human primate bipedalism analysis
Participants: Charles Pontonnier, Franck Multon, Armel Cretual.
A graphical representation of the baboon foot including joints axes. a visual representation of the ground reaction forces predicted with the method.
In this work, we have adapted an effective optimisation based ground reaction forces estimation method to the specific case of non-human primate bidepalism. Such optimisation seek at finding the minimal interaction forces guaranteeing the dynamical equilibrium of the subject against time. Studying Olive baboon, we developed a specific foot model with an additional flexion degree and a specific biomechanical model of the baboon. The method was evaluated on selected gait cycles, showing very conssitent results with previous ground reaction forces measurements. Published in 21.
8.3.2 Human-Exoskeleton interaction
The team is deeply involved in the evaluation of the biomechanical impact of assistive devices such as exoskeletons.
Operational and Biomechanical impact of a wrist exoskeleton to assist meat cutting tasks
Participants: Charles Pontonnier, Aurelie Tomezzoli.
See caption
Although a growing number of exoskeletons have been developed for occupational applications, wrist exoskeletons remain relatively rare. However, in the meat processing industry, elbow and hand-wrist musculoskeletal disorders are particularly common. The aim of this study was to assess the potential effectiveness and risks of a 670g wrist exoskeleton prototype designed to assist operators during meat cutting tasks. Six professional butchers performed three standardized tasks reproducing meat cutting gestures in foam, in three randomized experimental conditions: (1) without exoskeleton, (2) wearing the exoskeleton passive, with brakes off and (3) using it with brakes activated, locked in a static position. Cutting forces were recorded using an instrumented table, joint angles using an optoelectronic motion capture system, muscle activity using surface EMG and user experience was assessed using questionnaires. The cutting inaccuracy was defined as the area between the prescribed task and the actual cut on the foam surface. Joint torques were estimated by inverse dynamics with and without taking the exoskeleton’s mass into account, to isolate its effect. Linear mixed-effects statistical models were fitted. With the exoskeleton active during tasks, EMG activity was decreased of up to 18.7 % (p<0.01 to p<0.001) in the wrist flexors and increased of up to 61.7% (not significant to p<0.05) in the upper trapezius. Shoulder elevation joint torques were increased of up to 39.7% (p<0.001), mainly due to the exoskeleton mass. The proposed multi-criteria exoskeleton evaluation has provided guidance for following prototyping stages. Too heavy wrist exoskeletons could increase the risk of shoulder tendinitis for such tasks. Published in 31.
Mechanical caracterization of a passive upper limb exoskeleton for load carrying
Participants: Charles Pontonnier, Etienne Ricard.
Torque-angle curves obtained from an isokinetic ergometer
Adjusting the assistive torque of upper limb occupational exoskeletons is essential to optimize their effectiveness and user acceptance in companies. This adjustment enables a balance to be struck between the expected benefits and potential undesirable effects associated with their use, particularly for the shoulder joint, which is sensitive to the balance of forces. Despite this, no study has yet evaluated these assistive torques in static and dynamic conditions representative of work situations. The aim of this article is therefore to evaluate these assistive torques under these two conditions, using an isokinetic dynamometer. Angular velocities ranging from 0 to 240°/s and four levels of assistance were investigated. The results showed that the maximum assistive torques in flexion (energy restitution phase) were lower than those in extension (tensioning phase) by 20 to 36% and were median in static conditions. It was also observed that the level of assistance and the exoskeleton opening angles had a strong impact on the assistive torques, unlike the angular velocity in dynamic conditions, which had a minimal effect. Quantifying these assistive torques is crucial for assessing their biomechanical impact and adjusting the exoskeleton's assistance to the operator and the task performed. Published in 27. Part of this work has also been presented in 52, a preliminary study showing positive and negative effects of the assistive torque on shoulder and back biomechanics.
Passive upper limb exoskeleton based on a new spring design
Participants: Nicolas Vignais.
Upper limb exoskeletons may be valuable tools to decrease the risk to develop musculoskeletal disorders, and their design may differ from one manufacturer to another. The aim of the present study is to evaluate a passive upper limb exoskeleton based on a new spring design comprised of composite spring rods. Twelve right-handed participants have performed three manual tasks: static overhead holding, manual handling, and load carrying, with and without the upper limb exoskeleton. EMG activity of the anterior deltoid and biceps brachii was significantly reduced across the three tasks when wearing the exoskeleton. These findings suggested that the upper limb exoskeleton based on this new spring device permitted to reduce shoulder-flexor muscular demand through the tested experimental tasks. Longer-duration, in-field studies are now required to assess the efficacy of this exoskeleton for workers and potential implications for MSD prevention. Published in 17.
Upper limb movement regression using EMGs
Participants: Nicolas Vignais.
Participants were connected to the exoskeleton via the orthosis represented in green; interaction forces are measured via the F/T sensor represented in red. During the tasks, the yellow and white dots represented the current and target positions, respectively.
Achieving a versatile control of exoskeletons or prostheses requires accurate, robust and fast predictions of upcoming human movements. Electromyographic (EMG) signals are increasingly being used for this purpose as they theoretically allow to estimate joint torques in advance due to the electromechanical delay. However, their performance in continuously predicting joint torques strongly depends on the location of the sensors and on the feature extracted from the signal, whose impact on the accuracy is little-known in this context. In the present paper, the influence of different EMG features and muscles on the performance to predict both multi- and single-joint torques have been analyzed, while moving in a parasagittal plane hindered by a viscous force field applied by an exoskeleton. The existence of an accuracy–latency tradeoff is first highlighted, whether with frequency or time-domain features, the latter providing the best tradeoff. Second, the impact of muscle selection on the quality of the joint torques prediction is illustrated. In particular, the anterior deltoid, the long triceps, the brachialis, and the brachioradialis muscles are shown to be the most important to record during movements in a parasagittal plane. The present results pave the path towards the design of reactive and accurate exoskeletons and prostheses controllers based on EMG signals. Published in 24.
EMG-to-torque models for exoskeleton assistance
Participants: Nicolas Vignais.
In the field of robotic exoskeleton control, it is critical to accurately predict the intention of the user. While surface electromyography (EMG) holds the potential for such precision, current limitations arise from the absence of robust EMG-to-torque model calibration procedures and a universally accepted model. This paper introduces a practical framework for calibrating and evaluating EMG-to-torque models, accompanied by a novel nonlinear model. The framework includes an in situ procedure that involves generating calibration trajectories and subsequently evaluating them using standardized criteria. A comprehensive assessment on a dataset with 17 participants, encompassing single-joint and multi-joint conditions, suggests that the novel model outperforms the others in terms of accuracy while conserving computational efficiency. This contribution introduces an efficient model and establishes a versatile framework for EMG-to-torque model calibration and evaluation, complemented by a dataset made available. This further lays the groundwork for future advancements in EMG-based exoskeleton control and human intent detection. Published in 25 and a short communicatio about this work here 58.
Using vibratory feedback through an exoskeleton to generate ergonomic postures
Participants: Nicolas Vignais.
A picture with the setup and the corresponding trajectories.
In the past decades, active exoskeletons have been dedicated to reducing human effort, in particular to assist workers in occupational environments. However, this approach does not promote the learning of more ergonomic postures by workers, which is critical for the long-term prevention of musculoskeletal disorders. Alternatively, we propose the use of exoskeletons as biofeedback systems, generating task-relevant perturbations guiding users towards ergonomic postures. To test this approach, participants performed reach-to-hold movements towards a redundant target, allowing multiple final postures. We then introduced vibrations with posture-dependent intensity, generating a sensorimotor disturbance that canceled out either above or below each participant’s nominal preferred posture. Interestingly, participants adapted to minimize the vibrations, whether it increased or decreased the gravity efforts, and retained the novel posture when it induced lower effort. Finally, all participants significantly reduced effort post-exposure. This work demonstrates the feasibility of using exoskeletons as biofeedback systems to improve posture, paving the path for applications in musculoskeletal disorders prevention. Published in 14.
Effort generation capabilites mapping for personalized robotic assistance of the elbow
Participants: Mael Gallois, Nicolas Vignais, Charles Pontonnier.
A block scheme including the evaluation of the participant forces to adjust the level of assistance.
This study explores the use of a robotic aid to improve the mobility of people with neuromuscular or neurodegenerative disorders, who have reduced motor skills. It aims to compare the use of either a musculoskeletal model or a torque-angle joint model in a shared control framework of an upper limb exoskeleton in order to assists elbow flexion/extension movements. The method consists of measuring the effort-generating capacities of the elbow using an isokinetic ergometer. Data were then used to calibrate the two models. Both were integrated into an exoskeleton control law to estimate online user isometric capacities during a mass-holding task performed in various positions. Four experimental conditions were compared: without exoskeleton, with exoskeleton without assistance, with assistance computed from the quadratic model and from the musculoskeletal model. In order to test the method against individual variability, the study was conducted with nine participants. The results show a decrease in muscular activity for the biceps in both assistance conditions. In extrapolating effort-generating capacities outside the ergometer's measured range, the musculoskeletal model is more robust than the torque-angle joint model. Its estimations are more consistent with the literature, while torque-angle joint model errors induces unintended triceps activation. In conclusion, this study highlights the potential of combining a musculoskeletal model with shared control for assisting elbow. Published in38 and shorts communications here 13, 49.
8.4 Axis 4 : Ready-to-go field analysis
8.4.1 Computer vision
Exploiting Latent Properties to Optimize Neural Codecs
Participants: Pierre Hellier.
End-to-end image and video codecs are becoming increasingly competitive, compared to traditional compression techniques that have been developed through decades of manual engineering efforts. These trainable codecs have many advantages over traditional techniques, such as their straightforward adaptation to perceptual distortion metrics and high performance in specific fields thanks to their learning ability. However, current state-of-the-art neural codecs do not fully exploit the benefits of vector quantization and the existence of the entropy gradient in decoding devices. In this paper, we propose to leverage these two properties (vector quantization and entropy gradient) to improve the performance of off-the-shelf codecs. Firstly, we demonstrate that using non-uniform scalar quantization cannot improve performance over uniform quantization. We thus suggest using predefined optimal uniform vector quantization to improve performance. Secondly, we show that the entropy gradient, available at the decoder, is correlated with the reconstruction error gradient, which is not available at the decoder. We therefore use the former as a proxy to enhance compression performance. Our experimental results show that these approaches save between 1 to 3% of the rate for the same quality across various pretrained methods. In addition, the entropy gradient based solution improves traditional codec performance significantly as well. Published in 6.
D-Garment: Physics-Conditioned Latent Diffusion for Dynamic Garment Deformations
Participants: Pierre Hellier.
Adjusting and deforming 3D garments to body shapes, body motion, and cloth material is an important problem in virtual and augmented reality. Applications are numerous, ranging from virtual change rooms to the entertainment and gaming industry. This problem is challenging as garment dynamics influence geometric details such as wrinkling patterns, which depend on physical input including the wearer's body shape and motion, as well as cloth material features. Existing work studies learning-based modeling techniques to generate garment deformations from example data, and physics-inspired simulators to generate realistic garment dynamics. We propose here a learning-based approach trained on data generated with a physics-based simulator. Compared to prior work, our 3D generative model learns garment deformations for loose cloth geometry, especially for large deformations and dynamic wrinkles driven by body motion and cloth material. Furthermore, the model can be efficiently fitted to observations captured using vision sensors. We propose to leverage the capability of diffusion models to learn fine-scale detail: we model the 3D garment in a 2D parameter space, and learn a latent diffusion model using this representation independent from the mesh resolution. This allows to condition global and local geometric information with body and material information. We quantitatively and qualitatively evaluate our method on both simulated data and data captured with a multi-view acquisition platform. Compared to strong baselines, our method is more accurate in terms of Chamfer distance. Published in 59.
Generative Articulated Neural Fields to multiple identities and poses
Participants: Pierre Hellier, Franck Multon, Guillaume Loranchet.
See caption
The creation of realistic avatars in motion has been the hot-topic topic in academia and creative industry. Recent advances in deep learning and implicit representations have opened new avenues of research, especially to enhance the details of the avatars. However, these approaches generally lead to design large deep learning architectures dedicated to a specific identity shape and pose. Some previous works proposed to adapt the method to handle various poses for a given identity, but they are trained only on one identity. In VR applications, training this type of network for a new user, with a new shape, would require collecting many 3D geometric examples of this user in various poses. Moreover, for distant and massively populated virtual environments, it would also require to transfer one network for each user, which would require huge memory and network usage. This paper proposes an improvement over the state-of-the-art implicit SNARF and Fast-SNARF methods to permit generalization to novel motions and shape identities. Our result show that conditioning a light version of Fast-SNARF with a compact identity latent code enables to handle multiple shape identities while preserving the accuracy of the geometric reconstruction. Extrapolation to identity shapes that have never been seen during training is also possible, but with a little decrease of the reconstruction quality. Hence, to use his/her avatar in VR, a user simply has to provide to the application a compact identity latent code (obtained on a unique 3D example). Published in 51.
NIPIG: Neural Implicit Avatar Conditioned on Human Pose, Identity and Gender
Participants: Pierre Hellier, Franck Multon, Guillaume Loranchet.
See caption
The creation of realistic avatars in motion is a hot-topic in academia and the creative industries. Recent advances in deep learning and implicit representations have opened new avenues of research, especially to enhance the details of the avatars using implicit methods based on neural networks. State-of-the-art implicit Fast-SNARF [Chen et al. 2023] methods encodes various poses of a given identity, but are specialized for that single identity. This paper proposes NIPIG, a method that extends Fast-SNARF model to handle multiple and novel identities. Our main contribution is to condition the model on identity and pose features, such as an identity code, a gender indicator, and a weight estimate. Extensive experiments led us to a compact model capable of interpolating and extrapolating between training identities. We test several conditioning techniques and network's sizes to find the best trade-off between parameter count and result quality. We also propose an efficient fine-tuning approach to handle new out-of-distribution identities, while avoiding decreasing the reconstruction performance for in-distribution identities. Pubished in 41.
Implicit Shape Avatar Generalization across Pose and Identity
Participants: Pierre Hellier, Franck Multon, Guillaume Loranchet.
The creation of realistic animated avatars has become a hot-topic in both academia and the creative industry. Recent advancements in deep learning and implicit representations have opened new research avenues, particularly in enhancing avatar details with lightweight models. This paper introduces an improvement over the state-of-the-art implicit Fast-SNARF method to permit generalization to novel motions and shape identities. Fast-SNARF trains two networks: an occupancy network to predict the shape of a character in canonical space, and a Linear Blend Skinning network to deform it into arbitrary poses. However, it requires a separated model for each subject. We extend this work by conditioning both networks on an identity parameter, enabling a single model to generalize across multiple identities, without increasing the model's size, compared to Fast-SNARF. Published in 42.
8.4.2 Markerless motion analysis
Impact of introducing sparse inertial Measurement Units in Computer Vision-Based Motion Capture Systems for Ergonomic Postural Assessment
Participants: Charles Pontonnier, Franck Multon, Georges Dumont.
In ergonomics, worker movement on site is an important factor in assessing the risk of musculoskeletal disorders, among other factors. Several commercial markerless motion capture systems that can be used for this purpose are available, mostly based on monocular or multi RGB (THEIA system) / RGB-D cameras (MS Kinect system). Hybrid systems combining computer vision and Inertial Measurement Units (IMUs) have been introduced, such as the KIMEA (1 RGB-D+4 IMUs) and the KIMEA Cloud (1 RGB+4 IMUs) solutions. Although previous works analysed the accuracy of some of these systems, the relevance of coupling computer vision and IMU has not been studied. Hence, we tested the performance of these systems in evaluating bimanual handling tasks, with partial occlusions of the body in the images. The THEIA system exhibits an average of 11.1° error for all the joints, with larger Root Mean Square errors on the wrists and the shoulder (>14° error). KIMEA Cloud with IMU obtained similar global RMS error (10.3 ° to 10.9° depending on the viewpoint), but with better results for the wrists (3.9 ° to 4.3°). The impact of coupling RGB-D images and IMU data is even bigger: the RMS error of the Kinect decreased from 17.2° down to 8.9° when adding the IMUs information (KIMEA system). This difference is even bigger for the wrists: 28.3° to 38.5° for the Kinect, and 3.8 ° to 4° for KIMEA. These results confirm the advantage of introducing a few IMU sensors, especially for the wrists which are badly tracked in the images. Published in 33, 18.
Uncertainty estimation in marker-based motion capture of the tennis serve
Participants: Caroline Martin, Simon Ozan.
See caption
Marker-based motion capture (MoCap) systems are widely used to analyse human movement. However, they are affected by measurement uncertainties, particularly marker placement errors (MPE) and soft tissue artefacts (STA). Here, we quantify the individual and combined effects of these two sources of uncertainty on joint angles and angular velocities in the case of the tennis serve. A Monte-Carlo approach was used to simulate 3000 perturbed marker trajectories for each uncertainty source and their combination. We applied a random offset for MPE, while sinusoidal perturbations were used to simulate for STA. The resulting joint kinematics were compared across all degrees of freedom. Confidence intervals (5-95 %), root mean square deviation (RMSD) and Minimal Detectable Changes (MDC) were calculated for key biomechanical variables. Results showed that STA predominantly affected angular velocities, while MPE had a greater impact on joint angles. The combined simulation consistently produced the largest variability, with mean confidence intervals ranging from 5.1° to 30.8° for joint angles and from 70.5°.s-1 to 248.5°.s-1 for joint angular velocities, and RMSD values ranging from 1.6° to 8.4° for joint angles and from 16.8°.s-1 to 68.0°.s-1 for joint angular velocities. To our knowledge, this is the first quantification of MPE and STA effects on ballistic movement kinematics. These results provide critical reference values, enabling more accurate comparisons across subjects and studies while accounting for measurement uncertainties.Published in 20.
8.4.3 Tennis motion analysis and biomechanics
Tennis Serve Constraints Delimit but Do Not Prohibit Individual Movement Strategies among Professional Players
Participants: Caroline Martin, Simon Ozan.
The curve show a decrease of the com during knee flexion while being maximal at the impact of the ball
In sport, copying the motion technique of top-level athletes, which is regarded as perfect and optimal, is a goal often targeted by coaches for their athletes. However, inter-individual variability has been shown to remain in sport motions even at the elite level, and its analysis is important to identify the different motor strategies used to respond to the same motor task. This study aimed to assess inter-individual variability of Center of Mass (CoM) vertical kinematics during the serve among a homogeneous population of twenty-two professional tennis players and to investigate its links with serve performance indicators. Vertical CoM trajectories, impact height and racket head velocity were analyzed using a marker-based motion capture system. Inter-individual variability was assessed using the coefficient of variation. The phase duration, the vertical CoM displacement and its maximal velocity showed greater variability during the loading phase compared to the acceleration phase. Strong correlations were found between vertical CoM parameters during the acceleration phase and ball impact height, but not during the loading phase. The serving task constraints seem to offer limited scope for inter-individual variability and tend to delimit CoM movement strategy during the acceleration phase. Conversely, individual strategies of CoM kinematics differ among professional players during the loading phase, which could be the result of the expression of organismic constraints. This inter-individual variability observed during the loading phase points towards a non-unique way of serving and the need for individualized coaching exercises. Published in 10.
Contextual cues and stance decision-making in tennis forehand professional tennis players
Participants: Nicolas Vignais.
In tennis, both open and square stances are used to produce forehands, although few differences have been found between them in terms of biomechanical or ball speed advantages. Hence, it is currently unclear why expert players decide to change their stance during rallies. The objective of this study was to analyze the influence of the context before the stroke on the players’ stance choice. Based on video analysis of men and women professional tennis matches, 2479 stances chosen by twenty players were recorded, considering contextual factors such as “balance of power” and direction of the ball to be played. Results revealed that open stance was used significantly more than square stance in 83.6% of the forehands, with no difference between women and men. These rates varied based on the balance of power, with fewer open stance during favorable situations (65.4%). The direction of the ball played influenced the choice of stance in favorable situations, with fewer open stance being used to play cross court. Taken all together, these results showed an interaction between factors preceding the forehand stroke when adopting a stance in tennis. When players had little time to make a choice, they overwhelmingly used open stances. This choice can be automated, which reduces the attentional load. On the other hand, when they had more time, they varied their type of stance according to the zone being played, with inter-individual differences. The findings could provide valuable insights for coaches to optimize training and improve players’ decision-making efficiency. Published in 32.
Work in progress
Two preliminaries studies have been presented in congress regarding the reliability and usability of markerless motion capture for running 26 and the developent of a markerless method to reconstruct racket and ball motion during tennis serve 19.
9 Bilateral contracts and grants with industry
9.1 Bilateral contracts with industry
9.1.1 Nisk.AI, Industrial "defi" with Interdigital (2024-2030, partner)
Participants: Pierre Hellier, Leo-Paul Huar.
Nisk.AI (2024-2030) is a project within the Nemo joint laboratory INRIA-InterDigital. It is concerned with the efficient and frugal compression of new media. More specifically, we are concerned in Combo with the reconstruction and compression of 3D scenes using Gaussian Splats. This topic is managed through the PhD of Leo-Paul Huar.
9.1.2 Ys.AI, Industrial "defi" with Interdigital (leader)
Participants: Pierre Hellier, Franck Multon, Guillaume Loranchet.
Ys.AI (2020-2026) is a joint project with InterDigital aiming at exploring the representation of avatars in Metaverse environments. More specifically, we aim at pushing the limits of the uncanny valley for highly realistic avatars. To this end, we explore how to enhance fullbody, garments and hair simulation using AI approaches. We also explore how these methods could enhance the interaction experience in immersive worlds, with multisensory feedback rendering.
9.1.3 FFT, collaboration contract
Participants: Caroline Martin, Simon Ozan.
Contract with FFT (2023-2025)
This project was conducted in collaboration with the research and optimisation unit of the French Tennis Federation (FFT) to provide a biomechanical analysis of the serve for FFT players in a laboratory setting. The project was funded by the FFT to the tune of €22,400 over a period of 24 months. Caroline Martin was responsible for this project on behalf of the M2S laboratory at the University of Rennes 2.
9.1.4 VRboxing, collaboration contract
Participants: Richard Kulpa, Franck Multon, Benoit Bideau, Guillaume Claude, Julian Joseph, Annabelle Limballe.
Following the REVEA project, a collaboration was established between Université Rennes 2, Inria, the company VRBoxing, and the French Boxing Federation. The primary objective is to work collectively to better analyze combat sport situations and to develop innovative training protocols based on studies conducted in ecological settings. The second objective is to make these new defensive training approaches accessible not only to the French national boxing team but also to boxing clubs and amateur boxers.
9.2 Bilateral Grants with Industry
9.2.1 PerFoot
Participants: Anthony Sorel, Nicolas Bideau.
CIFRE with Stade Rennais (2023-2027)
This project, in collaboration with Stade Rennais FC, utilizes retrospective data and mixed models to chart the evolution of athletic qualities in young elite soccer players, factoring in age and prior injuries to better estimate future potential. By moving beyond static performance comparisons, the research accounts for longitudinal progress and individual variability, addressing a significant gap where most current methods fail to predict future capacity. This approach adapts statistical tools successfully used in skiing to a non-timed sport like soccer, aiming to provide a reliable evaluation system. Ultimately, the project targets key sporting, financial, and social objectives, such as optimizing investment and improving player career management.
9.2.2 VirGoal
Participants: Anthony Sorel, Nicolas Bideau.
CIFRE with Direction Innovation et Recherche of the French Football Federation (FFF) (2023-2026)
This project, in collaboration with the French Football Federation, focuses on developing and implementing virtual reality environments to enhance the perception-action coupling and decision-making capabilities of goalkeepers in football. The research pursues three primary objectives: 1) to identify the spatiotemporal variables and underlying paradigms governing goalkeeper decision-making in critical game situations, particularly the "back-pass" scenario; 2) to design and validate an ecologically valid virtual reality training application targeting perceptual-cognitive competencies; 3) to quantify the transfer of skills learned in the virtual environment to real-world pitch performance. The project's distinctive contribution lies in isolating and training often-unconscious decision factors in expert players through contextually-grounded virtual simulations, leveraging state-of-the-art immersive technologies and biomechanical analysis tools to bridge the gap between laboratory training and field performance.
9.2.3 Motor preferences and tennis
Participants: Caroline Martin.
CIFRE with Volodalen (2022-2025).
This project was carried out in collaboration with Volodalen in support of the ANR STHP BEST Tennis project and helped fund Kaies Deghaies' thesis on optimising tennis performance: biomechanical analysis of the serve and natural preferences.
9.2.4 BiomeCAP
Participants: Nicolas Vignais, Remy Roinson.
CIFRE with Phyling company (2023-2026).
The aim of this project is to detect risk of injury in running directly in the field by using embedded sensors. To this aim, the first step is to validate a running track instrumented with vertical force sensors. The second step is to develop a biomechanical model integrating kinematic data from a video-based markerless motion capture system and force data from the instrumented track. Finally, this methodology will be employed in the field to detect the risk of injury based on joint moment estimation.
9.2.5 SafranRV
Participants: Nicolas Vignais, Julie Robillard.
CIFRE with Safran Landing System company (2024-2027).
The aim of this project is to develop a virtual reality methodology to prevent the appearance of musculoskeletal disorders (MSDs) for aeronautic workers. To this aim, the first step is to record the real activity of these workers on a painting workstation using inertial sensor units, in collaboration with Safran LS company. These measurements will permit to obtain the reference activity. The second step is to model the workstation into a virtual environment, and to record the workers’ activity into this virtual situation. This will enable us to identify which are the critical elements (stereovision, haptics, etc.) to consider when using virtual reality for MSDs prevention. Finally, this virtual environment will be optimized and will be used to analyze the effect of familiarization in virtual reality on the task performance.
10 Partnerships and cooperations
10.1 European initiatives
10.1.1 Horizon Europe
Sharespace, (2023 – 2026, partner)
Participants: Valentin Ramel, Richard Kulpa, Julian Joseph, Benoit Bideau, Nicolas Vignais, Franck Multon, Anthony Sorel.
SHARESPACE will demonstrate a radically new technology for promoting ethical and social interaction in eXtended Reality (XR) Shared Hybrid Spaces (SHS), anchored in human sensorimotor communication. Our core concept is to identify and segment social sensorimotor primitives and reconstruct them in hybrid settings to build continuous, embodied, and rich human- avatar experiences. To achieve this, three interconnected science-towards-technology breakthroughs will be delivered: novel computational cognitive architectures, a unique self-calibrating body sensor network, and a fully mobile spatial Augmented Reality (AR) and virtual human rendering. We will create a library of social motion primitives and use them to design AI-based architectures of our artificial agents. SHARESPACE mobile capturing technologies combine loosely-coupled visual-inertial tracking of full body kinematic, hand pose and facial expression, incorporating novel neural encoding/decoding functionalities, together with local context-aware animations and highly realistic neural rendering. Our technology will be iteratively tested in 2 Proofs-of-principles involving human and artificial agents interacting in SHS, and 3 real-world use case scenarios in Health, Sport and Art. We will demonstrate a fully functional prototype of SHARESPACE tailored to the agents’ personalized characteristics (gender, culture, and social dispositions). SHARESPACE will support community-building and exploitation with concrete initiatives, including (i) public engagement around our research and innovation, (ii) promoting high-tech innovation and early transfer to our deep-tech companies, as premises for the consolidation of human-centric and sovereign European market areas such Industry AR and SHS, eHealth and tele-Health. Our long-term vision is to bring XR to a radically new level of presence and sociality by reconstructing sensorimotor primitives that enable ethical, trusted and inclusive modes of social interaction.
10.2 National initiatives
10.2.1 Continuum, ANR Equipex+
Participants: Georges Dumont, Ronan Gaugne, Anthony Sorel, Richard Kulpa.
Immersia and Immermove are the two parts of Immerstar platform led by Georges Dumont, In 2020, we participated to a PIA3-Equipex+ proposal. This proposal CONTINUUM involves 22 partners, has been succesfully evaluated and will be granted. The CONTINUUM project will create a collaborative research infrastructure of 30 platforms located throughout France, to advance interdisciplinary research based on interaction between computer science and the human and social sciences. Thanks to CONTINUUM, 37 research teams will develop cutting-edge research programs focusing on visualization, immersion, interaction and collaboration, as well as on human perception, cognition and behaviour in virtual/augmented reality, with potential impact on societal issues. CONTINUUM enables a paradigm shift in the way we perceive, interact, and collaborate with complex digital data and digital worlds by putting humans at the center of the data processing workflows. The project will empower scientists, engineers and industry users with a highly interconnected network of high-performance visualization and immersive platforms to observe, manipulate, understand and share digital data, real-time multi-scale simulations, and virtual or augmented experiences. All platforms will feature facilities for remote collaboration with other platforms, as well as mobile equipment that can be lent to users to facilitate onboarding. The kick-off meeting of continuum has been held in 2022, January the 14th. Global meetings are organized once a year, and the scientific committee meets every 2 or 3 months. Continuum network is part of French national strategy on research infrastructures since 2020 and applying to the renewal of this recognition this year.
10.2.2 AeX MusMaps, (2024 – 2028, co-leader)
Participants: Charles Pontonnier, Mael Gallois, Aurelie Tomezzoli, Nolwenn Fougeron, Nicolas Vignais, Elyse Larribau.
MusMapS uses effort measurements (mapping of physical abilities) from patients with disabilities to personalise the assistance provided by an upper limb exoskeleton with shared control, assisting these patients with everyday movements. The combination of a musculoskeletal model representing the patient's physical capabilities with shared control for conditions such as multiple sclerosis or stroke sequels is innovative and forms the core of the approach. Approx 250k funded by Inria.
10.2.3 PPR REVEA, (2020 – 2025, leader)
Participants: Richard Kulpa, Franck Multon, Benoit Bideau, Guillaume Claude, Julian Joseph, Annabelle Limballe.
PIA PPR Sport Très Haute Performance: The REVEA project (2020-2025) proposes a new generation of innovative and complementary training methods and tools to increase the number of medals at the Paris 2024 Olympic Games, using virtual reality. Indeed, the latter offers standardization, reproducibility and control features that: 1) Densify and vary training for very high performance without increasing the associated physical loads, and by IRISA Activity Report 2024 28 Inria Annual Report 2024 reducing the risk of impact and/or high intensity exercises ; 2) offer injured athletes the opportunity to continue training during their recovery period, or for all athletes during periods of confinement as experienced with Covid-19 ; 3) provide objective and quantified assessment of athlete performance and progress; and 4) provide a wide range of training that allows for better retention of learning and adaptability of athletes. Virtual reality offers a range of stimuli that go beyond the limits of reality, such as facing an opponent with extraordinary abilities or seeing an action that has not yet been mastered. The objective of REVEA is therefore to meet the needs of three federations by exploiting the unique properties of virtual reality to improve the motor performance of athletes through the optimisation of the underlying perceptual-motor and cognitive-motor processes. The French Gymnastics Federation wishes to optimise the movements of its gymnasts by observing their own motor production to avoid further increasing the load of physical training. The French Boxing Federation wishes to improve the perceptual-motor anticipation capacities of boxers in opposition situations while reducing the impact and therefore the risk of injury. The French Athletics Federation wishes to improve the perceptual-motor anticipation capacities of athletes in cooperative situations (4x100m relay) without running at high intensity. It is performed by a multidisciplinary consortium composed of University Rennes 2 (and Inria), University of Reims Champagne-Ardenne, Aix-Marseille University, Paris-Saclay University and INSEP.
10.2.4 PPR Neptune, (2021 – 2025, partner)
Participants: Guillaume Nicolas, Nicolas Bideau, Kilian Bertholon, Benoit Bideau, Antoine Bouvet.
PIA PPR Sport de Très Haute Performance. The objective of the NePTUNE project is to provide the French coaches and swimmers with quantitative tools and methods (1) to evaluate the key performance factors (stroke rate/stroke length ratio, velocity, water resistance, force, gliding efficiency, motor coordination, etc), (2) to identify the different performance profiles in order to individualize training, (3) to perform physical models in order to optimize performance. Thus, one of the issues developed in the framework of this project concerns the optimization of movement in world class swimming and in particular the links between hydrodynamic resistances and swimming technique. Based on innovative approaches, we propose to evaluate the previous key factors in experimental situations, from which knowledge can be transferred to coaches for the monitoring of swimmers. To achieve these objectives, we propose a research focus mixing biomechanics and fluid mechanics.
10.2.5 Thurston VR, CNRS 80 PRIME
Participants: Ronan Gaugne.
Name: The project aims to develop a virtual reality tool that allows users to explore non-Euclidean geometries in three dimensions in real time. These geometries, discovered in the 19th century by Gauss, Bolyai and Lobachevsky, challenge Euclid's fifth postulate and open up new mathematical “worlds”. The work of Thurston and Perelman shows that any three-dimensional space can be described by eight model geometries, known as Thurston geometries. The project seeks to simulate the visual perception of an observer in each of these spaces, where light does not necessarily follow straight paths. It is based on an existing web application and aims to enhance it and bring it to the Immersia immersive platform.
10.2.6 ANR MiXSON, (2024 – 2028, partner)
Participants: Caroline Martin, Simon Ozan.
Using an interdisciplinary approach, the MixSon project involves designing a sonification tool and examining its effectiveness and acceptance in a training context to optimise performance in a sporting movement: the tennis serve. This project brings together a specialist in the biomechanics of high-level sports movements, a neuroscientist who is an expert in movement sonification, an acoustician, a psychologist specialising in the acceptability of new technologies, and an anthropologist specialising in the analysis of the experiences of high-level athletes and coaches. I am responsible for WP1, which focuses on identifying the kinematic variables of the serve that are most relevant for sonification. The project, led by Guillaume Escalié (PU, University of Bordeaux), will last 48 months (total funding of €490.4K from the ANR).
10.2.7 Morphotwin
Participants: Nolwenn Fougeron.
CNRS MITI: Morpho Twin - Jumeaux numériques biomécaniques pour l’exploration des locomotions bipèdes humaines et non-humaines. Normalité et évolution
The MorphoTwin project aims in the creation of biomechanical digital twins of lower limbs, both human and primate, to unlock new insights into the structure and evolution of bipedal movement. By building diverse cohorts of 1,000 human and 20 non-human primate digital models, the team is capturing anatomical and functional. A groundbreaking robotic test bench allows researchers to validate these digital twins by simulating real-world locomotion on cadaver limbs. Early progress includes scans of primates and initial human data, with partnerships across hospitals and research centres to expand the dataset. Ultimately, MorphoTwin seeks to provide insight into human bipedalism and inform future medical, ergonomic, and evolutionary research.
10.2.8 Participation to PEPR ENSEMBLE
Participants: Anthony Sorel, Nicolas Vignais, Franck Multon, Richard Kulpa, Ahmed Abdourahman Mahamoud, Clelia Boulati, Ronan Gaugne.
eNSEMBLE (2023-2030) is an ambitious national project funded by the ANR PIA4 PEPR call. The eNSEMBLE project (Future of Digital Collaboration) aims to fundamentally redefine digital tools for collaboration. Whether it is to reduce the number of people on the move, improve territorial networking, or tackle the problems and transformations of the coming decades, the challenges of the 21st century will require collaboration at an unprecedented speed and scale. For this to happen, a paradigm shift in the design of collaborative systems is needed, comparable to the one that saw the advent of personal computing. This means inventing shared digital spaces that do more than simply replicate the physical world in virtual environments, enabling co-located and/or geographically distributed teams to work together fluidly and efficiently. In this context, ComBO is involved in the PhD thesis of Ahmed Abdourahman Mahamoud. The PhD topic consists in designing an AI-based controller of autonomous virtual humans that are supposed to behave as real human would do when interacting with users. More specifically, we explore imitation learning methods to train a controller to imitate the behavior of real humans in complex interaction tasks. ComBO is also involved in the PhD thesis of Clelia Boulati , started in December 2024. This PhD thesis is performed in collaboration with Racing92 rugby pro club, in Paris. It aims at developing a method based on virtual environments and intelligent systems to optimize the performance of players involved in touch-throwing in rugby. Last, ComBO is involved in two PhD theses in collaboration with the Seamless Inria team.The fisrt one ,' Shared Immersive Environments for Artistic Co-Creation in Heterogeneous Modalities', aims to design and validate new shared immersive environments that enable artists from various disciplines to collaborate in heterogeneous virtual spaces. The state of the art shows that current applications are mostly single-user and offer only limited forms of collaboration, rarely interdisciplinary. An initial prototype has been developed in Unity to facilitate interaction between dancers and musicians in a shared virtual environment. The dancer's movements, captured by motion tracking, can be replayed and used by the musician to modulate or create music based on their physical properties. The second, 'Nouvelles métaphores pour les performances collaboratives immersives' deals with new forms of artistic creation and performance, while raising scientific questions about 3D interaction, scenography and audience experience. These technologies make it possible to design musical or participatory performances in virtual environments shared between artists and spectators. The scenarios envisaged include immersive orchestras distributed in real time, as well as co-creation devices between performers and connected audiences. The main issues concern synchronisation, demonstration metaphors, multi-modal expression interfaces and inclusive collaborative tools. The overall objective is to adapt and design immersive metaphors that promote expressiveness, time management and the integration of the audience's role within immersive musical performances.
10.2.9 Participation to the PEPR ICCARE
GLAD-X project, leader
Participants: Charles Pontonnier, Ronan Gaugne.
The GLAD-X project recreates an immersive experience of Roman gladiatorial combat through a faithful virtual environment based on archaeological and biomechanical data. Users learn the moves of novice gladiators through avatars animated by motion capture. The experience will be enhanced by sound (shouts, ancient music), smell (sweat, blood) and touch (weapon vibrations). These additions aim to improve the sense of presence, historical accuracy and learning. Interaction with tangible interfaces (sword, shield) will enhance immersion. This device, which combines virtual reality, experimental archaeology and biomechanics, will be evaluated historically and from a user experience perspective. GLAD-X is fully in line with the objectives of the COMET programme on immersive multisensory content. 20k, funded by ICCARE.
10.2.10 Collaboration with "Institut Français du cheval et de l'équitation" and "fonds éperon"
EPICC PhD thesis
Participants: Nicolas Bideau, Benoit Bideau, Hugo Kerhervé, Nolwenn Regnault.
Equestrian performance relies on optimizing the rider–horse interaction (RHI), which is based on sensory information and postural control. In show jumping (SJ), the rider must accurately perceive distances in order to regulate the horse’s stride and successfully clear the obstacle. These perceptual–motor abilities are essential, as any deficit can lead to loss of confidence, increased stress, or injuries within the rider–horse pair. The EPICC project aims to identify the determinants of performance in show jumping through the study of rider–horse interactions, levels of expertise, and inter-individual differences. Three actions are undertaken: comparing expert and amateur riders, analyzing the effects of sensory perturbations, and testing perceptual training programs. The project seeks to define reliable indicators of rider–horse interaction and to improve training methods. Ultimately, EPICC aims to enhance performance, safety, and the well-being of the rider–horse pair.
HOPE PhD thesis
Participants: Hugo Kerhervé, Nicolas Vignais.
There is currently no consensus regarding best practices for training sport horses, with very few sport-specific metrics to assess performance and health besides veterinary science. This project aims to develop and validate non-invasive, field-specific tools for quantifying training load in young (2-3 yr) sport horses. Using these tools and borrowing from training methodologies developed in humans, a second objective is to evaluate the effect of different training strategies (continuous vs interval training) on performance and health indices.
11 Dissemination
11.1 Promoting scientific activities
11.1.1 Scientific events: organisation
General chair, scientific chair
- Charles Pontonnier : Co-organizer with Pascal Madeleine of a Symposium « human-system biomechanical simulation for ergonomics» at PREMUS 2025 ( 500 participants, work related musculoskeletal disorders prevention)
-
Franck Multon
:
- Exhibits and Sponsors Chair of IEEE VR2025
- Organizer of the Next Generation Avatar NGA2025 Workshop (IEEE VR2025 workshop)
-
Richard Kulpa
:
- Co-organizer of the 2nd DIGISPORT Winter School entitled “Quantify, Analyze, Prescribe: The Digital Shift in Physical Activity and Health”, November 2025
- Co-organizer symposium « XR en mouvement », June 2025
-
Ronan Gaugne
:
- Chair of the VR Tour of IEEE VR2025
- Organizer of the ARCHERIX IEEE VR Workshop
Member of the organizing committees
- Anthony Sorel : co-organizer of IEEE VR Tour during IEEEVR2025
11.1.2 Scientific events: selection
Member of the conference program committees
- Charles Pontonnier : Member of the scientific comitee of the 50th congrès de la Société de Biomécanique
- Diane Haering : Member of the comitee of the Prix Jean Vivès (Société de Biomécanique)
Reviewer
- Charles Pontonnier : Reviewer for the the 50th congrès de la Société de Biomécanique, Humanoids 2025, IROS 2025
- Franck Multon : Reviewer for ACM SIGGRAPH Asia2025, ACM SIGGRAPH 2025, ACM SIGGRAPH/Eurographics SCA2025, ACM Multimedia2025
- Nicolas Vignais : Reviewer for EuroXR 2025
- Nolwenn Fougeron : Reviewer for French Congress of Mechanics 2025
- Pierre Hellier : Reviewer for Eurographics 2025, PacificGraphics 2025
11.1.3 Journal
Member of the editorial boards
- Caroline Martin : Member of the editorial board of Journal of Medicine and Science in Tennis
- Charles Pontonnier : Associate editor of the episciences journal Multidisciplinary Biomechanics Journal (in charge of the track human system biomechanical interaction, robotics and ergonomics)
- Hugo Kerhervé : Member of the editorial board of PeerJ
- Mathieu Ménard : Member of the editorial board of International Journal of Osteopathic Medecine
- Nicolas Vignais : Member of the editorial board of Applied Ergonomics since 2019
Reviewer - reviewing activities
- Caroline Martin : Reviewer for British Journal of Sports Medicine, European Journal of Sports Sciences, Frontiers in Sports and Active Living, International Journal of Sports Physiology and and Coaching, Sports biomechanics
- Armel Cretual : Reviewer for Archive of Physical Medicine and Rehabilitation, Clinical Biomechanics, Disability and Rehabilitation, Human Movement Science, IEEE Journal of Engineering in Medicine and Biology, Journal of Motor Behavior, Medical and Biological Engineering and Computing, Musculoskeletal Science and Practice, Plos One, Scientific Reports
- Charles Pontonnier : Reviewer for IEEE Transactions on Visualization and Computer Graphics, Applied Ergonomics, Multibody Systems Dynamics, IEEE Computer Graphics and Applications, ACM Computing surveys, IEEE Robotics an Automation Letters, Journal of Biomechanics, Medical Engineering and Physics, Computer Methods in Biomechanics and Biomedical Engineering: Imaging and Visualization, Scientific Reports, Computer and Graphics, International Journal of Computer Assisted Radiology and Surgery, Helyon
- Franck Multon : Reviewer for Cognitive Development (Elsevier), Advanced Intelligent Discovery (Elsevier), Pattern Recognition (Elsevier), Pattern Recognition Letters (Elsevier), Information Fusion (Elsevier), Signal, Image and Video Processing (SpringerNature)
- Guillaume Nicolas : Reviewer for Journal of Biomechanics, Biophysics AIM, Sensors, Frontiers and active living, Journal of Human Kinetics
- Hugo Kerhervé : Reviewer for PeerJ, Physiological Reports, PLOS One, Clinical Physiology and Functional Imaging, Experimental Physiology, Journal of Sport Sciences, International Journal of Sports Medicine, European Journal of Sport Science, European Journal of Applied Physiology, Journal of Sport Science and Medicine, Journal of Athletic Enhancement, Human Factors and Ergonomics in Manufacturing and Service Industries
- Mathieu Ménard : Reviewer for Reviewer for International Journal of Osteopathic Medecine
- Nicolas Bideau : Reviewer for Sports Biomechanics, Reviewer for Journal of Sports Sciences
- Nicolas Vignais : Reviewer for Advanced Engineering Informatics, Ergonomics
- Nolwenn Fougeron : Reviewer for Biomedical Materials and Devices, Clinical Biomechanics, International Wound Journal
- Pierre Hellier : Reviewer for Signal, Image and Video Processing, Computer Animation and Virtual Worlds
11.1.4 Invited talks
- Anthony Sorel : "Virtual reality and elite sport: new frontiers in performance evaluation and training" presented in the “ International Summer School - Virtual Reality in Health and Sport”, Université de Caen Normandie, August 2025. Link
-
Caroline Martin
:
- “Exploring tennis environment to improve serve biomechanics in young players”, The 2025 ITF World Coaches Conference Overview, October 2025
- “How to improve the serve leg drive performance?”, The 2025 ITF World Coaches Conference Overview, October 2025
- “The biomechanics of open and neutral stances in tennis forehand: implications for performance and injury”, Tennis Australia Grand Slam Coaches’ Conference, January 2025
- “Tennis serve biomechanics: implications for performance and injury”, National Coaches Summit Tour Tennis Australia, January 2025.
-
Frank Multon
:
- “Défis et promesses de la réalité mixte pour l’entrainement sportif”, Colloque annuel du Centre d’Aide à la Performance Sportive CAPS, Liege, Belgium, November 2025
- “Physics-based simulation of Multi-Agent interactions using deep imitation learning”, Workshop du PC3 Matching, PEPR eNSEMBLE, Paris, October 2025
- Nolwenn Fougeron : “Identifiability of soft tissue material properties and uniqueness of hyperelastic constitutive models”, 20th International Symposium on Computer Methods in Biomechanics and Biomedical Engineering – CMBBE, September 2025
-
Pierre Hellier
:
- November 22, 2025: cérémonie de remise des diplômes de l'ISTIC, Keynote "L'écosystème image à Rennes : histoire et perspectives"
- November 6, 2025: invited talk at european seminar "AI FOR RESEARCH and EDUCATION" on infrastructure for AI education
- May 22, 2025: invited talk at Google Could Summit, AI education with cloud services
- April 30, 2025: invited talk at Breizh Macro Club (economical sciences university, Rennes) on AI infrastructure for education
- April 28, 2025: invited talk at creative machines, "Contrôle et apprentissage en IA"
11.1.5 Leadership within the scientific community
- Caroline Martin : Member of CNU 74
-
Franck Multon
:
- Member of the Scientific Committee of the GDR CNRS “Sport et Activités Physiques”
- Co-animator of the “DataPerf” Goupe de Travail of the GDR CNRS “Sport et Activités Physiques"
- coordinator of the "CORPS" group "Coordination des Organisations et Réseaux pour la Performance et les Sciences du sport" composed of INSERM, CNRS, C3D STAPS, MNSHS, Sciences2024, INSEP, "Agence Nationale du Sport" (Ministery of sporrs), APHP, and MESRI(Ministery of research)
- Mathieu Ménard : Member of Osteopathy Europe (scientific committee), Member of the C2R (reeeducation and rehadaptation commission) of the SFETD (Société Francaise d’Etude et de Traitement de la Douleur)
- Nolwenn Fougeron : Member of the board of directors of the French-speaking Society of Biomechanics
11.1.6 Scientific expertise
- Charles Pontonnier : Expert for ANR AAPG 2025
- Guillaume Nicolas : Expert for ANRT
- Nicolas Vignais : Expert for ANRT
11.1.7 Research administration
-
Charles Pontonnier
:
- Member of the pedagogical committee of the EUR DIGISPORT
- Co-animator of the group "biomeca@inria", aimed at coordiating biomechancial research in the Inria institute
- Member of the scientific comitee of the EQUIPEX+ Continuum
- Anthony Sorel :Technical manager ot the Immermove Research Platform, Member of UFR Staps comitee
-
Franck Multon
:
- Scientific Delegate of the Inria de l’Université de Rennes center since July 2025
- Coordinator of the national Inria actions in Sports Sport at Inria
- Co-animator of the Inria-INSEP convention, renewed in 2025 for 4 years
- Co-director of the Nemo.AI joint lab between Inria and InterDigital (end December 2026)
- Member of the EUR Digisport Scientific commitee
-
Georges Dumont
:
- Head of the EQUIPEX+ Continuum project for the four involved institutions from Rennes (ENS Rennes, INSA Rennes, University of Rennes, University of Rennes 2), co-leader of the scientific committee and of the executive committee
- member of the scientific commitee of EUR DIGISPORT
- Hugo Kerhervé : Member of the special interest group “Running Sciences” of the ECSS
- Mathieu Ménard : Head of research at Institut d’Ostéopathie de Rennes Bretagne
-
Nicolas Bideau
:
- Head of the Sport Performance in the Digital Era resaerch axis of M2S Lab
- Member of Scientific comitee of the EUR DIGISPORT
-
Richard Kulpa
:
- - Leader of PIA EUR DIGISPORT project
- Member of the Academic and Scientific Committees of the University Rennes 2
- Head of the research theme “Efficient Extended Environments” of Inria ComBO team
11.2 Teaching - Supervision - Juries - Educational and pedagogical outreach
11.2.1 Teaching
-
Anthony Sorel
:
- 3D Performance Analysis, Master 1, 24h, Master STAPS, Univ Rennes 2
- 3D Performance Analysis, Master 2, 24h Master STAPS, Univ Rennes 2
-
Armel Cretual
:
- Biomécanique spécifique aux APAS, L2 and L3,30h, Licence APA-Santé, Univ Rennes 2
- Méthodologie de la recherche et métrologie, Master 1, 20h, Masters APA et IEAP, Univ Rennes 2
- Evaluation fonctionnelle, Master 2, 20h, Master APA, Univ Rennes 2
-
Caroline Martin
:
- Théorie spécialité tennis, Licence 1, 18h, Licence STAPS, Univ Rennes 2
- Pratique spécialité tennis, Licence 1, 18h, Licence STAPS, Univ Rennes 2
- Théorie spécialité tennis, Licence 2, 36h, Licence STAPS, Univ Rennes 2
- Pratique spécialité tennis, Licence 2, 36h, Licence STAPS, Univ Rennes 2
- Connaissances scientifiques et théoriques en spécialité tennis (biomécanique et physiologie), Licence 3, 84h, Licence STAPS, Univ Rennes 2
- Encadrement de stagiaires, M1 – M2, 20h, Master STAPS EOPS et DIGISPORT, Univ Rennes 2
- Préparation aux épreuves d’admission de l’agrégation externe d’EPS Prépa Agreg (3ème année ENS), 5h, ENS Rennes
-
Charles Pontonnier
:
- Mécanique Lagrangienne, stabilité et vibrations, Master 2, 40h, M2 ISC, ENS Rennes
- Cosimulation humain-systèmes, Master 2, 24h, M2 EEEA parcours SIVOS - Human Centered Robotics, Univ Rennes
- Modélisation biomécanique de l’humain, Master 2, 24h, M2 EEEA parcours SIVOS - Human Centered Robotics, Univ Rennes
- Systèmes d’assistance, master 2, 24h, M2 EEEA parcours SIVOS - Human Centered Robotics, Univ Rennes
-
Diane Haering
:
- Biomécanique, M1, 18h Master STAPS tronc commun, Univ Rennes 2
- Biomécanique, L2, 12h, Licence Entraînement Sportif, Univ Rennes 2
- Biomécanique, L2, 12h, Licence Ergonomie, Univ Rennes 2
- Mathématiques, L3, 6h, Licence Ergonomie, Univ Rennes 2
- Statistiques, M1-M2, 18h, Master STAPS mutualisé, Univ Rennes 2
- Exploration Fonctionnelle, M2, 16h, Master APPCM, Univ Rennes 2
- Biomécanique, L1, 138h, Licence STAPS tronc commun, Univ Rennes 2
- Gymnastique, L2, 3h, Licence Entraînement, Univ Rennes 2
- Energie et mouvement humain, M1, 3h, Master IEAP, Univ Rennes 2
-
Georges Dumont
- Mechanical simulation in Virtual reality, M2, 18h, Master Engineering of complex systems and Mechatronics, Rennes University and ENS de Rennes
- Mechanics of deformable systems, M2, 40h, Master ISC, ENS Rennes
- oral preparation to agregation competitive exam, M2, 24h, Master ISC, ENS Rennes
- Vibrations in Mechanics, M2, 12h, Master ISC, ENS Rennes
- Finite Element method, M2, 12h, Master isometric, ENS Rennes
- Guillaume Nicolas Biomechanics, Master – licence, 240h, Licence 1, Licence 2 STAPS, Licence ES STAPS, Master EOPS, Master SNS Digisport, Univ Rennes 2
-
Hugo Kerhervé
- Scientific basis of strength and conditioning, L3, 50h L3 sports training, Univ Rennes 2
- Testing in swimming, L3, 3h, L3 sports training, Univ Rennes 2
- Optimizing sports training methodologies, M1, 9h, Master sports training, Univ Rennes 2
- Adaptations in extreme conditions, M1, 18h, Master sports training, Univ Rennes 2
- Sports planification and programming, M1, 15h Master sports training, Univ Rennes 2
- Coupling physiology and biomechanics, M1, 6h, Master sports training, Univ Rennes 2
- Metrology, M1, 6h, Master sports training, Univ Rennes 2
- Professional development, M1, 12h, Master sports training, Univ Rennes 2
- Internships, M1, 20h, Master sports training, Univ Rennes 2
- Quantifying training load, M2, 30h, Master sports training, Univ Rennes 2
- Fatigue and recovery strategies, M2, 30h, Master sports training, Univ Rennes 2
- Internships, M2, 30h, Master sports training, Univ Rennes 2
-
Mathieu Ménard
:
- Biomechanics, M1, 6h, Master APAS, Univ Rennes 2
- Research methodology, M2, 14h, Master MEFRS, Univ Haute Alsase (SERFA)
-
Nicolas Vignais
:
- Ergo des postes et grilles, L2, 21h, Univ. Rennes 2
- conditions environnementales et travail,L3, 13h, Univ Rennes 2
- Tutorat et projet, L3, 6h, Univ Rennes 2
- Intervention and entrepreneuriat, M1, 6h, Univ Rennes 2
- Introduction aux données sensibles, M1, 9h, Univ Rennes 2
- Revue de littérature, M1, 4,5h, Univ Rennes 2
- Zotero et biblio, M1, 4,5h, Univ Rennes 2
- Journée Analyse in situ Gemba, M1, 3h, Univ Rennes 2
- projet entrepreneuriat, M2, 6h, Univ Rennes 2
- Capacité à concevoir des assistances technologiques, M2, 7h Univ Rennes 2
- séminaire thématique,M2, 3,5h, Univ Rennes 2
- cosimulation et mouvement humain, M2 SNS, 1,5h, Univ Rennes
- Ergonomics and chiropraxy, 5A, 3h, IFEC
- outils d'analyse et physiologie musculaire, L2, 6h Univ Rennes 2
- muscle et force : physio et bioméca, L2, 12h, Univ Rennes 2
- environnement physique, L2, 16h, Univ Rennes 2 outils et méthodes d'analyse en physiologie L2 12 Univ. Rennes 2
- environnement physique de travail, M1, 3,5h , Univ Rennes 2
- ergo et optimisation de la perf, M1, 7,5h, Univ Rennes 2
- Analyse in situ, M1, 4h, Univ Rennes 2
- projet entrepreneuriat - évaluation, M1, 6h, Univ. Rennes 2
- syst assistance, exo, acceptabilité, M1, 7,5h, Univ Rennes 2
- Recherche Développement Conception, M1, 8h, Univ Rennes 2
- Physical ergonomics and industrial performance, Master, 10h, INSA Rennes
-
Nolwenn Fougeron
:
- Human system cosimulation, Master, 3h, M2 SIVOS-HCR and M2 SNS, ENS Rennes
-
Pierre Hellier
:
- Optimisation continue, Master, 11h, Master IA, Univ Rennes
- Introduction à l’IA, L3, 50h, Licence info, Univ rennes
- Optimisation machine learning, Doctoral studies, 30h, Doctorat, UM6P (Maroc)
-
Ronan Gaugne
:
- Réalité virtuelle, Master, 18h, Master Création numérique, Univ Rennes 2
- Projet Info, Master, 10h, 4e année INFO, INSA Rennes
- Etudes Pratiques, L3, 22h, 3e année INFO, INSA Rennes
11.2.2 Supervision
- Arthur Guillotel (2023-2026):Approche pluridisciplinaire de la détection de talents et du passage entre haut potentiel et footballeur professionnel, supervised by Anthony Sorel by Brigitte Gelein (ENSAI) and Benoit Bideau , Univ Rennes 2
- Koffi Amezoui (2023-2026): Analysis and clustering of soccer game situations in order to populate virtual environments, supervised by Mathieu Marbac (ENSAI), Brigitte Gelein (ENSAI) and Anthony Sorel , ENSAI
- Florian Matéos (2023-2026) : Biofidélité de l’interaction pied-ballon en réalité virtuelle au service de l’évaluation et l’optimisation de la prise de décision au football, supervised by Anthony Sorel , Richard Kulpa and David Mann (VU Amsterdam University), Univ Rennes 2
- Mathis Tiercery (2023-2026) : Conception d’environnements virtuels pour l’analyse et l’entrainement des compétences spécifiques du gardien de but au football, supervised by Benoit Bideau ,Anthony Sorel , Univ Rennes 2
- Clelia Boulati (2024-2027): Analyse des coordinations spatio-temporelles en sport collectif pour la modélisation comportementale en environnement virtuel : application à la touche au rugby, supervised by Franck Multon , Nicolas Vignais and Anthony Sorel , Univ Rennes
- Nolwenn Poquerusse (2025-2028): Evaluation de l'agilité réactive des athlètes sur le terrain, supervised by Nicolas Vignais and Anthony Sorel , Univ Rennes 2
- Léo Denouel (2025-2028): Étude des relations entre l'activité musculaire, les modifications des tissus mous et les stratégies de mouvement chez des individus souffrant de lombalgie persistante afin de favoriser leur retour au sport et à l'activité physique, supervised by Mathieu Ménard , Diane Haering ,Nolwenn Fougeron and Armel Cretual , Univ Rennes 2
- Simon Ozan (defended 12/12/2025): Validation et exploitation d’un système de capture du mouvement markerless pour analyser et optimiser in situ la biomécanique du service de joueurs de tennis de haut niveau, supervised by Caroline Martin and Richard Kulpa , Univ Rennes 2
- Kaies Deghaies (2021-2026): Stratégies motrices et variabilité inter-individuelle du service au tennis chez les joueurs de tennis experts : vers la démystification d’un modèle technique idéal, supervised by Benoit Bideau , Thibault Lussiana and Caroline Martin , Univ Rennes 2
- Louis Arles (2024-2027): Comprendre et optimiser l’action des membres inférieurs dans le cadre du service au tennis : Analyse des déterminants biomécaniques et des liens avec la performance et les risques de blessures, supervised by Richard Kulpa and Caroline Martin , Univ Rennes 2
- Guillaume Le Guludec (2025-2028): Visual observations to Biomechanical quantities: building a differentiable visual and biomechanical model to retrieve biomechanical quantities from visual observations, supervised by Pierre Hellier , Laurent Albera, Caroline Martin and Charles Pontonnier , Univ Rennes
- Alexis Jensen (2022-2026): Simulation de foule dense par modélisation dynamique, supervised by Julien Pettré (Inria VirTUS) and Charles Pontonnier , Univ Rennes
- Etienne Ricard (2022-2026): Modélisation musculosquelettique du système « homme-exosquelette », supervised by Charles Pontonnier , Chris Hayot (INRS) and Kevin Desbrosses (INRS), ENS Rennes
- Aurelien Schuster (2023-2026): Modèle musculosquelettique du fantassin : vers une analyse énergétique de l'activité physique en mission pour l'optimisation des équipements et du chargement, supervised by Georges Dumont and Charles Pontonnier , ENS Rennes
- Suzon Pucheu (2024-2027): Quantification des impacts de réception sur agrès en gymnastique à partir de méthodes de localisation et d'estimation de la source, supervised by Charles Pontonnier , Diane Haering and Mathieu Aucejo (CNAM Paris), ENS Rennes
- Elisa Jolas (2024-2027): Analyse numérique de l’effet des orthèses de genou pour le maintien d’une activité physique adaptée tout au long de la vie, supervised by Charles Pontonnier , Raphael Dumas, Rachid Aissaoui, Sacha Guitteny, Univ Lyon 1
- Mael Gallois (2024-2027): Commande partagée par cartographie musculaire pour l’assistance du membre supérieur pour des pathologies neuromusculaires et neurodégénératives, supervised by Nicolas Vignais , Charles Pontonnier , Marie Babel and Sylvain Guégan, INSA Rennes
- Elyse Larribau (2025-2028): Conception et commande d'un exosquelette de poignet-main basé sur la modélisation de biomécanique personnalisée d'usagers en situation de handicap à mobilité résiduelle, supervised by Charles Pontonnier , Nolwenn Fougeron , Marie Babel and Sylvain Guégan, ENS Rennes
- Sony Saint-Auret (2022-2026): Virtual Collaborative « Jeu de Paume », supervised by Franck Multon , Richard Kulpa ,Ronan Gaugne and Valérie Gouranton, INSA Rennes
- Shubhendu Jena (Defended june 2025): Combining implicit and explicit representations for modeling 3D Shape and appearance, supervised by Franck Multon a nd Adnane Boukhayma, Univ Rennes
- Guillaume Loranchet (2023-2026): Deep interactive control of virtual character’s motion based on separating identity, motion and style, supervised by Franck Multon , Pierre Hellier , João Regateiro and Adnane Boukhayma, Univ Rennes
- Ahmed Abdourahman Mahamoud (2023-2026): MAIIL - AI-driven character simulation based on Multi-Agents Interaction Imitation Learning, supervised by Franck Multon , Richard Kulpa , Ewa Kijak and Simon Malinowski, Univ Rennes
- Victor Restrat (2023-2026): PAOLI: Pole Vault generic analysis, human motion and optimal interaction, supervised by Guillaume Nicolas , Nicolas Bideau and Georges Dumont
- Yannis Raineteau (defended, april 2025): Apport du profilage force-vitesse pour l’évaluation du transfert de la préparation physique à sec vers la performance en natation sprint de haut niveau, supervised by Guillaume Nicolas and Benoit Bideau , Univ Rennes 2
- Souebou Bouro (defended, december 2025): Multi-Sensor System for In-Situ Motion Analysis of Kayaker, supervised by Guillaume Nicolas , Nicolas Bideau , Olivier Berder and Antoine Courtay, Univ Rennes
- Dingye Zhang (2021-2026): Associations Between Land-Based and In-Water Lower-Body Strength in Swimming: A Load–Velocity Approach, supervised by Benoit Bideau , Nicolas Bideau and Guillaume Nicolas , Univ Rennes 2
- Titouan Perrin (Defended, 3/12/25): La restriction du flux sanguin : mécanismes vasculaires et musculaires sur la performance physique en escalade, supervised by Hugo Kerhervé , Mathieu Marillier (INSERM) and Julien Brugniaux (UGA), Univ Grenoble Alpes
- Romain Demay (2021-2026): La durabilité, facteur discriminant de la performance lors de l’exercice d’endurance intermittent : aspects fondamentaux et appliqués, supervised by Hugo Kerhervé and Carole Groussard (UR2), Univ Rennes 2
- Lola Masson (2023-2026): La relation entre raideur et économie de course : quelle place dans le modèle de performance en course à pied ? supervised by Hugo Kerhervé and Guillaume Millet (UJM), Univ Rennes 2
- Lisa Poitevineau (2024-2027): Monitoring de l'entraînement : évaluation de la performance et de l'état de forme du cheval de sport, supervised by Hugo Kerhervé and Anne Couroucé (Oniris), Univ Rennes 2
- Antoine Bouvet (Defended, May 2025) : Monitoring and Modeling Swimming Performance Using Inertial Measurement Units and Data Science Methods and Applications for Scientific Support in Performance Optimization, supervised by Nicolas Bideau and Mathieu Marbac (CREST-ENSAI), Univ Rennes 2
- Nolwenn Regnault (2023-2026): Perceptual Training for Riders to Optimize Rider–Horse Interactions in Show Jumping, supervised by Nicolas Bideau , Benoit Bideau and Agnès Olivier, Univ Rennes 2
- Kilian Bertholon : Underwater Motion Analysis Using Embedded Sensors: An Inertial Measurement Unit (IMU)–Based Approach for In-Situ 3D Motion Capture and Biomechanical Assessment of the Four Swimming Strokes, supervised by Benoit Bideau and Nicolas Bideau , Univ Rennes 2
- Thomas Chevalier (2022-2026): Méthodologie pour l'analyse biomécanique du sportif de haut niveau à partir de capteurs embarqués: Application au cyclisme, supervised by Nicolas Vignais and Laetitia Fradet (UBS), Univ Paris-Saclay
- Remy Roinson (2024-2027): Développement d'une méthodologie d'évaluation in situ des facteurs de risques de blessure en course à pied , supervised by Nicolas Vignais and Jean-Philippe Boucher (Phyling), Univ Rennes 2
- Julie Robillard (2024-2027): Développement d’une méthodologie basée sur les environnements virtuels de conception pour anticiper le risque de TMS en industrie ˸ influence de la présence et de l’immersion, supervised by Nicolas Vignais and Estelle Chin, Univ Rennes 2
- Clément Horteur (2022-2026): Intérêt de la planification et modélisation pré-opératoire sur ju-meau numérique pour le choix des critères d'implantation d'une prothèse totale de genou, supervised by Nolwenn Fougeron , Yohan Payan and Antoine Perrier, Univ Grenoble Alpes
- Ysé Roch (2024-2027): Génération d'une cohorte de jumeaux numériques du genou humain, évaluation des paramètres morpho-fonctionnels et simulation biomécanique. Application à la planification de pose de prothèse totale du genou, supervised by Nolwenn Fougeron , Yohan Payan and Antoine Perrier, Univ Grenoble Alpes
- Antoine Dumoulin (2023-2026): Video-based dynamic garment representation and synthesis, supervised by Pierre Hellier , Stefanie Wuhrer (Inria Morpheo), Joao Regateiro (InterDigital) and Adnane Boukhayma (Inria), Univ Greboble
- Leo-Paul Huar (2025-2028): Compression de représentations dynamiques de scènes 3D par approche de Gaussian splatting, supervised by Christine Guillemot (Inria), Neus Sabater (InterDigital), Gustavo Sandri (InterDigital) and Pierre Hellier , Univ Rennes
- Antoine Monier (2025-2028): Hybrid conventional and deep learning-based video coding, supervised by Aline Roumy (Inria), Fabrice Le Leannec (InterDigital), Karam Naser (InterDigital) and Pierre Hellier
- Theo Gerard (2025-2028): A multimodal Deep-Learning approach for intuitive editing of character animation, supervised by Pierre Hellier , Marc Christie (Univ Rennes) and Ludovic Hoyet (Inria), Univ Rennes
- Valentin Ramel (2023-2026): Perception-Action Dynamics and Synchronization in Extended Reality Peloton Cycling, supervised by Richard Kulpa and Benoit Bardy, Univ. Rennes 2
- Julien Lomet (Defended,Nov 2025): La collaboration dans une œuvre en réalité virtuelle : du processus de création à la performativité, supervised byCédric Pleissiet, Valérie Gouranton and ronan gaugne Univ Paris 8
- Maxime Dumonteil (2023-2026): Archéologie Cognitive en Environnement Virtuel, supervised by Ronan Gaugne , Valérie Gouranton, Marc Macé and Théophane Nicolas (Inrap), Univ Rennes
- Nathan Salin (2024-2027): Shared Immersive Environments for Artistic Co-Creation with Heterogeneous Modalities, supervised by Florent Berthaut (Univ Lille), Valérie Gouranton, and Ronan Gaugne , Univ Rennes
- Guillaume Vallet (2024-2027): Nouvelles métaphores pour les performances collaboratives immersivessupervised by Florent Berthaut (Univ Lille), Valérie Gouranton, and Ronan Gaugne , Univ Lille
- Maé Mavromatis (2025-2028): ThurstonVR, réalité virtuelle dans des espaces non-Euclidiens, supervised by Rémi Coulon (CNRS), Valérie Gouranton and Ronan Gaugne , INSA Rennes
11.2.3 Juries
PhD juries
-
Caroline Martin
:
- Nicholas Busuttil January 2025 « The use of physically-constraining tools for skill development: an investigation into a grip position training tool during common tennis strokes », Examiner
- Lucio Caprioli, November 2025, "New Systems for Functional Assessment in Sports Contexts: Use of Inertial Sensors in the Technical-Tactical Analysis on the Tennis Court", Examiner
- Thomas Daney, November 2025, "Interactions sportif-équipement dans le cas du padel : effet de la variabilité inter-raquette et de l’usure", Examiner
-
Charles Pontonnier
:
- Diana Pardo Ramos, Caractérisation biomécanique du travail de manutention dans le milieu de la petite enfance, Univ Lyon 1, 28/04/2025, Reviewer
- Aymeric Orhan, Exploitation de démonstrations artificielles et de la variabilité humaine pour la commande par apprentissage d’un exosquelette de membre supérieur, Université Paris-Saclay, 25/06/2025, Reviewer
- Maialen Matray, Modélisation du membre résiduel chez les personnes amputées d'un membre inférieur pour la conception d'emboitures personnalisées en fabrication additive, ENSAM ParisTech, 22/09/2025, Reviewer
- Benoit Hureaux, Synthèse de lois de commande bio-inspirées à impédance variable pour interaction avec l’environnement dans des tâches rythmiques, Université Paris-Saclay, 12/12/2025, Reviewer
- Franck Multon : Théo Cheynel, Institut Polytechnique de Paris, December 2025, Character Animation : Bridging the Gap Between AI Models and Artistic Animation Pipelines , President
-
Georges Dumont
:
- Renjie Zhang, 27/05/2025, Production system design and improvement methodology based on VR and human factor performance indicators, President
- Antoine Raud, 27/08/2025, Modélisation musculo-squelettique du tronc et des membres supérieurs : application à l’escrime fauteuil, Reviewer
-
Nicolas Bideau
:
- Damien Hello, October 2025, Approche pluridisciplinaire et connectée de la performance en natation et des représentations en 5ème nage, Examiner
- Pauline Algourdin, November 2025, Etude mécanique et biomécanique de l’interaction Homme / Palme : efficience and performance, Examiner
-
Nicolas Vignais
:
- Clément Thévenot, March 2025, Influence de l’utilisation du dispositif d’assistance physique CORFOR sur les contraintes articulaires et neuromusculaires lors d’une tâche de préparation de commandes manuelle du secteur agroalimentaire, Reviewer and President
- Yosra Tounekti, April 2025, Biomechanical characterization and modeling of human-machine interaction: personalized in situ prevention approach, Reviewer
- Kéliane Mégret, October 2025, Conception de matériaux architecturés multistables pour les interfaces automobiles, President
- Jason Dellai, October 2025, Évaluation de l’apprentissage d’un nouveau geste technique en faveur de la prévention des troubles musculo-squelettiques : application au secteur de la coiffure, Reviewer
- Florian Cazorla, December 2025, Identifier des facteurs de risque modifiables de cervicalgie chez les travailleurs et les déterminants de l’implémentation d’un programme de prévention, Reviewer
- Maxime Manzano, December 2025, Contributions à l’ergonomie et aux stratégies de commande des exosquelettes de suppléance du membre supérieur, President
-
Nolwenn Fougeron
:
- Tianyu Jia, March 2025, Finite element modelling of the mechanical functionality of medical devices made from biocompatible metastable beta titanium alloys, Examiner
- Mathieu Severyns, December 2025, Analyse biomécanique des ménisques et du cartilage, après chirurgie de réparation méniscale et procédure d’ostéotomie tibiale en chargement sous IRM 7T, Examiner
-
Pierre Hellier
:
- Aymen Merrouche, November 13 2026, Learning Non-Rigid 3D Surface Matching, Reviewer
- Tom Bordin, December 4 2026, Compression sémantique d’images à extrêmement bas débits, Examiner
-
Richard Kulpa
:
- Aymeric Erades,December 2025, Visual Analysis of Table Tennis Tactics, Reviewer
- Alexandre Schortgen, June 2025, Implementation, evaluation and application of markerless motion analysis for in situ performance assessment of elite athletes, Reviewer
- Alexis Laly, June 2025, Locomotion in intellectual developmental disorder and autism spectrum disorder: influence of a Mixed Reality intervention on spatiotemporal and kinematic gait parameters, Examiner
11.3 Popularization
11.3.1 Productions (articles, videos, podcasts, serious games, ...)
- Anthony Sorel : Article for “France Football » entitled « Quand la réalité virtuelle aide les gardiens de but à faire le bon choix », May 2025, Link
- Caroline Martin : France Info, « Wimbledon : le service, toujours une arme fatale sur gazon ? », july 2025
- Charles Pontonnier : Sortir la biomécanique du laboratoire, Emergences, 2025. Link.
- Nolwenn Fougeron : Podcast Biomécanique and Cie. Link.
11.3.2 Participation in Live events
- Anthony Sorel : Regular demos of Motion Capture and VR capabilities of the Immermove Research Platform as part of conferences organization (e.g. IEEE VR, ISMAR) or for visitors, whether scientist, industrials or politics
- Caroline Martin : Open de Rennes de tennis, roudn table « Data et Sport », Rennes, september 2025
-
Franck Multon
:
- “Défis et promesses de la réalité mixte pour l’entrainement sportif », SciRennes, Inria Rennes
- June 2025, Démo in a 4-days booth, invited at AI For Good (United Nations sponsored event), Geneva, July 2025
-
Pierre Hellier
:
- October 9, 2025: Dissemination to undergrad students at "fête de la science", GenAI for image generation and editing
- June 5, 2025: participation to "ciné-débat", dissemination to general public on AI and societal impact
-
Ronan Gaugne
:
- Démo at Recto VRso (Laval Virtual, Apr. 2025), Journées Européennes d’Archéologie (St Malo, Jun. 2025)
- Festival Court-Métrange (Rennes, Sept. 2025), Fête de la Science (Rennes, Oct. 2025)
12 Scientific production
12.1 Major publications
- 1 articleAccuracy evaluation of an automatic landmarking method based on 3D segmented CT scans of full lower limbs with knee osteoarthritis.European Journal of Radiology191October 2025, 112292HALDOI
- 2 inproceedingsNIPIG: Neural Implicit Avatar Conditioned on Human Pose, Identity and Gender.CVMP 2025 - 22nd ACM SIGGRAPH European Conference on Visual Media ProductionLondres, United KingdomACM2025, 1-15HALDOI
- 3 articleUncertainty estimation in marker-based motion capture of the tennis serve.Journal of Biomechanics1952025, 113119HALDOI
- 4 articleA size-adjustable instrumented track to measure vertical ground reaction forces: a validation study.IEEE Sensors Journal2510March 2025, 16720-16725HALDOI
- 5 articleOperational and biomechanical evaluation of a wrist exoskeleton prototype for assisting meat cutting tasks.IEEE Transactions on Human-Machine Systems2025, 1-10HALDOI
12.2 Publications of the year
International journals
International peer-reviewed conferences
Conferences without proceedings
Reports & preprints
Other scientific publications
Scientific popularization
12.3 Cited publications
- 66 articleVirtual reality to assess and train team ball sports performance: A scoping review.Journal of sports Sciences3822020, 192--205back to text
- 67 articleBiomechanical fidelity of athletic training using virtual reality head-mounted display: the case of preplanned and unplanned sidestepping.Sports Biomechanics2022, 1--22back to text
- 68 articleDesign and verification of a human--robot interaction system for upper limb exoskeleton rehabilitation.Medical engineering & physics792020, 19--25back to text
- 69 articleNeural and biomechanical tradeoffs associated with human-exoskeleton interactions.Applied ergonomics962021, 103494back to text