Section: Overall Objectives

Statement

Object recognition —or, in a broader sense, scene understanding— is the ultimate scientific challenge of computer vision: After 40 years of research, robustly identifying the familiar objects (chair, person, pet), scene categories (beach, forest, office), and activity patterns (conversation, dance, picnic) depicted in family pictures, news segments, or feature films is still beyond the capabilities of today's vision systems. On the other hand, truly successful object recognition and scene understanding technology will have a broad impact in application domains as varied as defense, entertainment, health care, human-computer interaction, image retrieval and data mining, industrial and personal robotics, manufacturing, scientific image analysis, surveillance and security, and transportation.

Despite the limitations of today's scene understanding technology, tremendous progress has been accomplished in the past ten years, due in part to the formulation of object recognition as a statistical pattern matching problem. The emphasis is in general on the features defining the patterns and on the algorithms used to learn and recognize them, rather than on the representation of object, scene, and activity categories, or the integrated interpretation of the various scene elements. WILLOW complements this approach with an ambitious research program explicitly addressing the representational issues involved in object recognition and, more generally, scene understanding.

Concretely, our objective is to develop geometric, physical, and statistical models for all components of the image interpretation process, including illumination, materials, objects, scenes, and human activities. These models will be used to tackle fundamental scientific challenges such as three-dimensional (3D) object and scene modeling, analysis, and retrieval; human activity capture and classification; and category-level object and scene recognition. They will also support applications with high scientific, societal, and/or economic impact in domains such as quantitative image analysis in science and humanities; film post-production and special effects; and video annotation, interpretation, and retrieval. Machine learning is a key part of our effort, with a balance of practical work in support of computer vision application and methodological research aimed at developing effective algorithms and architectures.

WILLOW was created in 2007: It was recognized as an Inria team in January 2007, and as an official project-team in June 2007. WILLOW is a joint research team between Inria Paris, Ecole Normale Supérieure (ENS) and Centre National de la Recherche Scientifique (CNRS).

This year we have hired Justin Carpentier as an Inria researcher. Jean-Paul Laumond has joined Willow as a senior CNRS researcher. Both Justin and Jean-Paul have robotics background and will strengthen the team in its new efforts towards learning embodied representations and robotics. Vladimir Petrik and Pierre-Yves Masse have been visiting post-docs within the framework of collaboration with the Intelligent Machine Perception project lead by J. Sivic at the Czech Technical University in Prague. Hazel Doughty has been a visiting PhD student from University of Bristol. Minttu Alakuijala, Aamr El Kazdadi, Pierre-Louis Guhur and Bruno Lecouat have joined Willow as new PhD students.