-
reconstruction Develop methods for behavior understanding from multi-view video and audio recordings Advance multimodal large language models for fine-grained action understanding and reasoning Help shape the next
-
to redefine how machines interpret human actions, anticipate intentions, and interact seamlessly with people in various settings. The successful candidate will be developing new methods for Real-time Video
-
the restitution workshops held in year 4. Contribute to the production of the Vertical Talks video capsules and podcasts, together with the project team and the local partners. Scientific production Publish at
-
Postal Code 1700 STATUS: EXPIRED X (formerly Twitter) Facebook LinkedIn Whatsapp More share options E-mail Pocket Viadeo Gmail Weibo Blogger Qzone YahooMail
-
this new hybrid network. Your code will directly enable the next generation of energy-efficient AI clusters. Project scope You will bridge the gap between custom optical hardware and standard AI software
-
methodological expertise. Contact details for three professional references. Optional but encouraged: links to code repositories, analysis pipelines, experimental software, technical projects or other outputs