The different research activities in the cluster share the same general objectives aiming at automatically understanding multimedia data (images , video , digital documents, 3D scenes). They focus on acquisition/reconstruction, indexing , modeling, classification or automatic content recognition (objects, actions , concepts ). The concept of "visual object" at the heart of our work is a common denominator of our research.

Skills gathered in the cluster

The research activities in the cluster propose gateways to cross the semantic gap between low -level descriptors contained in raw images and representations of higher semantic levels (modeling, classification and identification). The researchers of the cluster have recognized skills on signal processing (filtering, segmentation, feature extraction), machine learning and pattern recognition (connectionist, statistical and structural approaches) , information fusion , constraint programming, discrete and continuous optimization.

Application fields, expertise at the interface of other disciplines

The research conducted in the cluster covers a wide range of methodologies to provide solutions to application areas related to the automatic analysis of multimedia data. Among these applications, without being exhaustive , we can cite the analysis of written documents (automatic structuring, search by content and enrichment), the detection/recognition of visual objects or concepts including the analysis of 2D/3D faces and the recognition of actions and movements in the videos.

Researchers of the cluster participate to numerous collaborations with teams of different disciplines, which, beyond the Sciences and Technologies of Information and Communication, also concern the Humanities and Social Sciences and the Sciences of Ecology and Environment.

Teams and researchers of the cluster

The 16 permanent researchers participating in the cluster belong to the teams:

• IMAGINE (16)

Keywords

• Video and Image processing
• Computer Vision
• Pattern Recognition
• Machine Learning
• Image Representation
• Scene Analysis
• Design and analysis of algorithms
• Artificial Intelligence