About this talk
This talk, presented by Axel Farrugia and Iryna de Albuquerque Silva from CEA, introduces Eclipse Aidge, an open-source platform designed to simplify the deployment of AI models on edge devices. The session covers how Eclipse Aidge addresses the fragmentation in edge AI toolchains, enabling seamless interoperability across various machine learning frameworks, compilers, and hardware SDKs. Key features include a modular architecture that supports the ONNX format, transparent intermediate representations for model manipulation, as well as optimisation techniques like quantisation and tensor compression. The platform generates portable C++ code for deployment on diverse hardware architectures, ensuring that visibility and control are maintained throughout the AI model lifecycle.
