PhD on Low Power Image Sensor for Distributed Processing in Cameras Network

This PhD position focuses on designing innovative, low-power image sensors and readout architectures to enable distributed artificial intelligence processing within wireless camera networks. Hosted at CEA-Leti's L3i Laboratory in collaboration with Université Grenoble Alpes, the project seeks to overcome the energy limitations of traditional centralized cloud architectures by implementing analog ASIC inference cores directly at the sensor nodes. The successful candidate will define, design, test, and fabricate a demonstrator that integrates image acquisition with local AI pre-processing for IoT applications. This role requires working closely with academic and industrial partners using professional integrated circuit design tools and software development environments.

• Define, design, and test an innovative readout architecture for standard image sensors to extract strategic features efficiently.
• Collaborate with academic partners to establish overall mote architecture, including data formats and output procedures.
• Develop analog-based ASIC inference cores capable of performing local AI pre-processing computations.
• Integrate low-power hardware solutions to support distributed inference and image acquisition in wireless sensor networks.
• Contribute to the realization and testing of an image sensor hardware demonstrator using professional IC design tools.