Design of power- and area-efficient spiking neural network for a high density neural recording system
Type and duration:
Master project, flexible duration in collaboration with IMEC The Netherlands
Context:
Neural recording systems play a pivotal role in understanding the intricate workings of the brain and unlocking its mysteries. By merging computational neuroscience and hardware optimization, we seek to unlock groundbreaking insights into brain function while advancing next-generation neural recording technologies.
Objectives:
The project aims at designing a power- and area-efficient spiking neural network to perform rapid feature detection by employing the spatial information from different channels, for closed loop neuromodulation.
Required skills:
- Knowledge of neural networks
- Basic knowledge of circuit or RTL design
- Fundamental understanding of neural systems
Contact person:
Yuming He
References:
- Y. He et al., "An Implantable Neuromorphic Sensing System Featuring Near-Sensor Computation and Send-on-Delta Transmission for Wireless Neural Sensing of Peripheral Nerves," in IEEE Journal of Solid-State Circuits, vol. 57, no. 10, pp. 3058-3070, Oct. 2022, doi: 10.1109/JSSC.2022.3193846.
- J. Stuijt et al., "μBrain: An event-driven and fully synthesizable architecture for spiking neural networks." Frontiers in neuroscience 15 (2021): 664208.
Last modified: | 31 July 2023 3.31 p.m. |