Section: New Results

People Detection, Tracking and Re-identification Through a Video Camera Network

Participants : Malik Souded, François Brémond.

keywords: People detection, Object tracking, People re-identification, Region covariance descriptors, SIFT descriptor, LogitBoost, Particle filters.

This works aims at proposing a whole framework for people detection, tracking and re-identification through camera networks. Three main constraints have guided this work: high performances, real-time processing and genericity of the proposed methods (minimal human interaction/parametrization). This work is divided into three separate but dependent tasks:

People detection:

The proposed approach optimizes state-of-the-art methods [89] , [93] which are based on training cascades of classifiers using the LogitBoost algorithm on region covariance descriptors. The optimization consists in clustering negative data before the training step, and speeds up both the training and detection processes while improving the detection performance. This approach has been published this year in [46] . The evaluation results and examples of detection are shown in Figures 24 and 25 .

Object tracking:

The proposed object tracker uses a state-of-the-art background subtraction algorithm to initialize objects to track, with a collaboration of the proposed people detector in the case of people tracking. The object modelling is performed using SIFT features, detected and selected in a particular manner. The tracking process is performed at two levels: SIFT features are tracked using a specific particle filter, then object tracking is deduced from the tracked SIFT features using the proposed data association framework. A fast occlusion management is also proposed to achieve the object tracking process. The evaluation results are shown in Figure 26 .

People re-identification:

A state-of-the-art method for people re-identification [67] is used as a baseline and its performance has been improved. A fast method for image alignment for multiple-shot case is proposed first. Then, texture information is added to the computed visual signatures. A method for people visible side classification is also proposed. Camera calibration information is used to filter candidate people who do not match spatio-temporal constraints. Finally, an adaptive feature weighting method according to visible side classification concludes the improvement contributions. The evaluation results are shown in Figure 27 .

This work has been published in [28] .