This paper addresses the problem of probabilistic recognition of activities from local spatio-temporal appearance. Joint statistics of space-time filters are employed to define histograms which characterize the activities to be recognized. These histograms provide the joint probability density functions required for recognition using Bayes rule. The result is a technique for recognition of activities which is robust to partial occlusions as well as changes in illumination.In this paper the framework and background for this approach is first described. Then the family of spatio-temporal receptive fields used for characterizing activities is presented. This is followed by a review of probabilistic recognition of patterns from joint statistics of receptive field responses. The approach is validated with the results of experiments in the discrimination of persons walking in different directions, and the recognition of a simple set of hand gestures in an augmented reality scenario.

This paper addresses the problem of the local scale parameter selection for recognition techniques based on Gaussian derivatives. Patterns are described in a feature space of which each dimension is a scale and orientation normalized receptive field (a unit composed of normalized Gaussian-based filters). Scale invariance is obtained by automatic selection of an appropriate local scale [Lin98b] and followed by normalisation of the receptive field to the appropriate scale. Orientation invariance is obtained by the determination of the dominant local orientation and by steering the receptive fields to this orientation. Data is represented structurally in a feature space that is designed for the recognition of static object configurations. In this space an image is modeled by the vectorial representation of the receptive field responses at each pixel, forming a surface in the feature space. Recognition is achieved by measuring the distance between the vector of normalized receptive fields responses of an observed neighborhood and the surface point of the image model. The power of a scale equivariant feature space is validated by experimental results for point correspondences in images of different scales and the recognition of objects under different view points.

This paper presents a new technique for the perception and recognition of activities using statistical descriptions of their spatiotemporal properties. A set of motion energy receptive fields is designed in order to sample the power spectrum of a moving texture. Their structure relates to the spatio-temporal energy models of Adelson and Bergen where measures of local visual motion information are extracted by comparing the outputs of a triad of Gabor energy filters. Then the probability density function required for Bayes rule is estimated for each class of activity by computing multi-dimensional histograms from the outputs from the set of receptive fields. The perception of activities is achieved according to Bayes rule. The result at each instant of time is the map of the conditional probabilities that each pixel belongs to each one of the activities of the training set. Since activities are perceived over a short integration time, a temporal analysis of outputs is done using Hidden Markov Models. The approach is validated with experiments in the perception and recognition of activities of people walking in visual surveillance scenari. The presented work is in progress and preliminary results are encouraging, since recognition is robust to variations in illumination conditions, to partial occlusions and to changes in texture. It is shown that it constitute a powerful early vision tool for human behaviors analysis for smart-environnements.

No abstract