Sparse neural networks enable low-power, implantable neural interfaces
Joseph T. Costello, Hisham Temmar, Luis Hernan Cubillos, Matthew J. Mender, Dylan M Wallace, Madison M. Kelberman, Omer Benharush, Jeremy Simons, Matthew S. Willsey, Nishant Ganesh Kumar, Theodore A. Kung, Parag G. Patil, Cynthia A. Chestek
This paper shows that sparse neural networks reduce power consumption in implantable BMIs by pruning decoders, enabling efficient decoding with minimal performance loss.
bioRxiv · Dec 17, 2024 · PreprintBioengineering