Song Han is an assistant professor at MIT’s EECS. His research focuses on efficient deep learning computing and TinyML techniques, including model compression, hardware-aware neural architecture search, and AI accelerators.