After its reorganization at the beginning of 2017, the department is now organized into three groups:
- Signal, Statistics and mAchine learning- S2A – is a group that is devoted to the applications of statistics and probability to the field of information processing. The research area covers a wide spectrum from the development of new techniques and new algorithms to various applications. The activities of the group comprises the following topics: statistical learning, independent data and complex random systems, methods and algorithms for cosmological data analysis, Markov Chain Monte-Carlo techniques, sequential Monte-Carlo techniques (particle filters), array processing, geolocalization, models estimation. It also conducts activities in signal processing having strong connections with the physical phenomenon that is at the source of the signals, whether acoustical or coming from other sensors. In digital audio signal processing, the activities span the entire acquisition chain, from capture to analysis or transformation, transmission up to its restitution, with the goal of proposing solutions to the main problems centered around the sound, speech or music, in multimedia applications.
- Image, Modeling, Analysis, Geometry, Synthesis (IMAGES) – The IMAGES team (Image, Modeling, Analysis, GEometry and Synthesis) is active in the fields of image processing and analyzing, computer vision and computer graphics. Its contributions span a large scope and are concerned with modeling, analyzing, transforming, representing, interpreting and synthesizing images, tri-dimensional and numerical objects. The team focuses on the global solving of complex image and computer graphics problems, ranging from raw data processing to high-level interpretation. Its contributions draw from multiple and complementary methods, relying both on applied mathematics and computer science. The main research topics are: mathematical image models, artificial intelligence for image understanding and spatial reasoning, computer graphics for geometric modeling, image synthesis, virtual reality and 3D interactive systems. The team deals with applications in medical imaging, remote sensing imaging, computational photography and video games.
- Multimedia – MM- is a group that covers the life cycle of multimedia documents in the framework of a complete chain going from authoring tools for on-line and offline production of multimedia contents to multimodal interaction for the final user; this also includes automated processing like enhancement of degraded pictures, verification of the identity of the user, modification of auditive and visual appearance, image segmentation and pattern recognition. The group also works on techniques that allow the analysis, compression and robust transmission of these media in heterogeneous networks. It also works on the dynamic and distributed adaptation of the transmitted data flow (including meta-data and in particular those concerning the digital rights management) with respect to context, transport conditions and terminal type.