ANR Collaborative Research Project
PoInt Cloud compressioN for human perceptIon and maChine viSion
PICNICS explores next-generation point cloud compression methods that jointly consider human perceptual quality and machine vision performance, with the goal of enabling immersive communication, efficient storage, and AI-ready visual representations.
Visual quality, immersive media, perceptual optimization
Task-aware coding, semantic analysis, AI-ready representations
About
Point clouds are becoming a key representation for immersive media, three-dimensional scene understanding, robotics, and extended reality. Their efficient coding requires more than minimizing bitrate and signal distortion: future methods must also preserve the information needed by downstream machine analysis while maintaining high perceptual quality for human users.
PICNICS investigates coding frameworks, representations, and evaluation methodologies that bridge human-centric and task-oriented objectives. The project combines expertise in multimedia coding, visual quality assessment, machine vision, and learning-based methods.
Partners
PICNICS brings together three complementary academic partners across multimedia signal processing, quality assessment, and machine vision.
Laboratoire des Signaux et Systèmes
CNRS, CentraleSupélec, Université Paris-Saclay
Project coordinator
Partner investigator
Université de Poitiers
Partner investigator
Add a portrait or team image later.
Sorbonne Université
Partner investigator
Add a portrait or team image later.
News
Use Pages CMS to add the first news item.
Open Positions
Future PhD, postdoc, internship, or engineer openings can be posted here through Pages CMS.
Publications
Replace this placeholder with accepted journal and conference papers, preprints, datasets, or software releases.
You can later link PDFs, code, arXiv pages, or project deliverables.
Code & Resources
Code, datasets, and evaluation resources can be shared through your project repository as they become available.
Replace this link with the actual GitLab or GitHub repository when ready.
Contact
For collaborations, questions, or student opportunities related to PICNICS, feel free to contact the project coordinator.