Partners of Brain-Net
Brain-Net is funded by ANR over the period 2021-2024. The project gathers 3 academic partners with expertise in neurotechnology, neuroscience, neural computing, and brain-computer interfaces, as well as an industrial partner developing neuromorphic hardware for real-time pattern analysis.
The Neurotechnology and network dynamics group at the Grenoble Institue of Neuroscience investigates cortical networks underlying vocal production in mammals and speech in humans using large-scale cortical implants and aims at developing speech rehabilitation solutions using brain-computer interfaces. The group develops SNN architectures for real-time audio and neural data processing.
The CERCO has a strong expertise in training spiking neural networks, using unsupervised, supervised, or self-supervised learning. He will provide the SNN models able to decode and translate audio and brain signals, trained with data acquired by the GIN partner, and implementable on the FPGAs developed by IMS, or on the BrainChip Akida chip.
The Biolelectronic group at IMS laboratory, University of Bordeaux, develops innovative systems for health technology. The group especially designs real-time neuromorphic circuits like SNNs to mimic biological neural systems and to process neural information. Brainmorphic algorithms are created and implemented in embedded hardware such as FPGA.
Brainchip is revolutionizing Edge AI applications with an event domain neural processor, the Akida™ hardware and its accompanying comprehensive development environment. Akida AI IP, AI SoC and development environment help customers create ultra-low power solutions which are scalable and flexible. The custom solutions can achieve high performance while keeping networks small and power consumption ultra-low thus enabling a wide array of edge capabilities, including continuous learning and inference.