Analysis of spatio-temporal data
In this topic, we specifically address tasks related to (i) motion analysis where video sequences are processed to estimate velocity or track moving objects or particles, (ii) human expression and behavior understanding which includes detection and recognition of face expressions, micro-expressions, gestures and specific events or actions related to human activities. The originality of our approach is to process heterogeneous data as signal, image, depth, video...
Our aim is to develop new formalisms to represent, manipulate and visualize the information in challenging contexts as heterogeneous image data (color videos, sequences of depth images, thermal and spectral image sequences, ...) or a set of videos associated to multi-views of the same scene. The main objectives are to describe various motions, behaviours, expressions (a lot of contributions are still expected when the data are not acquired in a constrained context like a person facing a camera) or micro-expressions. We will have a particular focus at working at the crossroads of different approaches as probabilistic and deterministic modeling, unsupervised and supervised clustering, ...
The main challenges which can be met in various projects are how to detect, track and analyze multi-events in heterogeneous image data, which implies to manage properly the complementary information ; how to manage the quantity of data which can be huge ; ...
Various application domains like technologies for the health can be tackled. Hereafter are given examples of topics we addressed: