The Embed-AI project is part of a broader effort to design artificial intelligence systems that are both resource-efficient and high-performing. It is built around two complementary innovations. On one side, it relies on a decentralized network of small smart cameras, or motes, equipped with low-resolution sensors and onboard analog neural processors. These devices collaborate to process data directly at the source, minimizing the need for data transmission. This approach helps preserve privacy, reduces overall energy consumption, and enhances the robustness of the system.
On the other side, to overcome the inherent computational and energy limitations of such lightweight devices, the project is developing a specialized analog ASIC chip. This chip is designed to optimize distributed neural network operations and can handle complex tasks, such as matrix–vector multiplications, with extremely low power consumption.
By bringing these two elements together, Embed-AI paves the way for a sustainable, decentralized, and scalable model of AI. The system makes it possible to envision robust edge computing applications that process large amounts of data locally while keeping material and energy requirements to a minimum.
ACTIVITIES
Each call for position will be posted in websites and distributed though mailing lists. Open positions will be advertised with clear project descriptions. Individuals from under-represented categories will be encouraged to apply. There will be a pre-selection based on the evaluation of CVs. The short-listed candidates will be interviewed by the appointing PI. After that, the PI will report to the RTB. Gender equality policies will be applied in case of candidates with very similar qualifications.
For this chair, 4 positions are opened: 3PhD thesis and 1 postdoc.
CHAIR PRESENTATION
License:
Unless otherwise stated, all documents are shared under the Creative Commons BY-NC-ND 4.0 license.
You may view and share them for non-commercial purposes, without modification, and with appropriate credit to the authors.
Published on August 27, 2025 Updated on August 27, 2025
Core members
François BERRY
Omar AIT AIDER
Emmanuel BERGERET
Gilles SICARD
Laurent FESQUET
Stéphane MANCINI
Associated members
Maxime PELCAT
Ricardo CARMONA
Hassen AZIZA
Research topics
TinyML, Edge computing, AI on reconfigurable devices, Analog computing, Analog storage of weights, Synthesis of elementary operators
Share the linkCopyCopiedClose the modal windowShare the URL of this pageI recommend:Consultable at this address:La page sera alors accessible depuis votre menu "Mes favoris".Stop videoPlay videoMutePlay audioChat: A question? Chatbot Robo FabricaMatomo traffic statisticsX (formerly Twitter)