Yihui He

full stack AI research engineer🧠🤖

I'm a senior research engineer at Headroom, where I launched key AI features:
I was a research enginear at Facebook AI Research, where I worked on large scale video self-supervised learning, fairmotion, PyTorchVideo
2018 - 2019
MS in computer vision at Carnegie Mellon.
2014 - 2018
Majored in computer science at Xi'an Jiaotong University. During my undergrad study, I had a track record of contributing to CNN efficient inference.
  • I designed channel pruning to effectively prune channels. (2k citations)
  • I further proposed AutoML for Model Compression to sample the design space of channel pruning via reinforcement learning, which greatly improved the performance. (1k citations)
  • I served as a reviewer for ECCV'20, ICML'20, CVPR'20, ICLR'20, ICCV'19, CVPR'19, ICLR'19, NIPS'18, Pattern Recognition Letters, TIP and IJCV.
Featured projects
We motivate and present feature selective anchor-free (FSAF) module, a simple and effective building block forsingle-shot object detectors. It can be plugged into single-shot detectors with feature pyramid structure. The FSAF module addresses two limitations brought up by the conventional anchor-based detection: 1) heuristic-guided feature selection; 2) overlap-based anchor sampling.
We introduce a novel bounding box regression loss for learning bounding box transformation and localization variance together. The resulting localization variance is utilized in our new non-maximum suppression method to improve localization accuracy for object detection. On MS-COCO, we boost the AP of VGG-16 faster R-CNN from 23.6% to 29.1% with a single model and nearly no additional computational overhead. More importantly, our method improves the AP of ResNet-50 FPN fast R-CNN from 36.8% to 37.8%, which achieves state-of-the-art bounding box refinement result.
misc unsorted