Bias-aware neural architecture search
2025 – 2028
PI: Enzo Tartaglione
Investigates neural architecture design that accounts for social/statistical biases directly, aiming for fairer models by construction rather than post-hoc correction.
Projects
Funded research projects led or co-led by members of the Multimedia team
The team is involved in projects spanning compression, multimedia distribution, frugal AI, geometric and multimodal learning. Below is a snapshot of currently ongoing and recent projects. To add a new project, drop a Markdown file into this folder (see the template in this site’s Nextcloud share).
Ongoing
DeSNAP — Deep Simplicial Neural Networks for Advanced Geometry Processing
2025 – 2028
PI: Jhony H. Giraldo
Develops deep learning architectures based on simplicial complexes to model higher-order interactions in geometric and graph-structured data.
GPAC — open multimedia framework
2003 – present
PI: Jean Le Feuvre
Open-source multimedia framework for interactive content creation, packaging, distribution and playback — continuously funded through 25+ EU/French R&D projects and industry sponsorship since 2003.