Song Han
Assistant Professor at Massachusetts Institute of Technology
Cambridge, Massachusetts, United States
Summary
🤩
Rockstar🎓
Top SchoolSong Han is an assistant professor at MIT EECS and a leading researcher in efficient deep learning computing with 11 years of experience bridging algorithms and hardware. He pioneered "deep compression" and the Efficient Inference Engine to shrink and accelerate neural networks, work that earned Best Paper awards, an NSF CAREER Award, and recognition as one of MIT Technology Review’s “35 Innovators Under 35.” His hardware-aware neural architecture search and edge-AI toolchains have been integrated into frameworks like PyTorch and AutoGluon and repeatedly won low-power vision contests at CVPR, ICCV and NeurIPS. Trained at Tsinghua and Stanford, he combines theoretical rigor with practical system building—publishing widely while translating research into deployable accelerators for mobile and tiny-ML platforms. Based in Cambridge and active in both academia and industry collaborations, he focuses on making state-of-the-art models run efficiently on constrained devices.
11 years of coding experience
Bachelor of Science (B.S.), Electrical Engineering, Bachelor of Science (B.S.), Electrical Engineering at Tsinghua UniversityDoctor of Philosophy (PhD), Doctor of Philosophy (PhD) at Stanford University
Github Skills (82)
high-resolution10
nvidia-jetson-nano10
imagenet9
caffe9
model-compression9
pruning9
compression9
shift9
re-identification9
segmentation9
pointnet9
tinyml8
video-understanding8
vision-transformer8
optical-flow8
Programming languages (6)
Github contributions (5)
songhan/SqueezeNet-Generator
Mar 2016 - May 2018
Contributions:17 pushes, 1 branch in 2 years 2 months
songhan/Deep-Compression-AlexNet
Apr 2016 - Mar 2019
Deep Compression on AlexNet
Contributions:15 pushes, 2 branches in 2 years 11 months
pytorchcaffe2deep-learningcompressioncaffe
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.