News

• Our new paper "boundary thickness and robustness in learning models" is accepted by NeurIPS 2020! We show that a "thick" decision boundary helps improve robustness of neural networks against both adversarial examples and out-of-distribution transforms. 
• Our paper on applying coded computing techniques to non-von Neumann computing architectures is published in the Proceedings of IEEE!! It is encouraging that more and more people realize the importance of fault tolerance at the gate level, e.g., see my previous paper on this topic. 
• Our work FoldingNet was selected as a spotlight talk in CVPR! You can see a video here. The code and the paper are also released. It shows that transforming a regular grid into the shape of a point cloud can provide useful point cloud decoder with state-of-the-art unsupervised learning ability. We also have a concurrent paper in CVPR called KCNet here with code release. KCnet provides a novel way of exploiting neighborhood structures in point clouds using graph-based and kernel-correlation-based techniques.

• Research Intern at the TLC group, Microsoft, mentored by Dr. Matteo Interlandi and Dr. Saeed Amizadeh, summer 2018
• Research Intern at MERL, mentored by Dr. Chen Feng and Dr. Dong Tian, summer 2017
• Research Intern at Bell Labs, mentored by Dr. Mohammad Ali Maddah-Ali, summer 2016
• Undergraduate thesis research in Department of Electronic Engineering, Tsinghua University, advised by Professor Wei Chen and Professor Bo Bai, 2013
• Research assistant in Institute for Interdisciplinary Information Sciences, Tsinghua University, advised by Professor Longbo Huang, 2013
• Research Intern in UC Davis, mentored by Professor Qing Zhao, summer 2012
• Student research training in Department of Electronic Engineering, Tsinghua University, mentored by Professor Pingyi Fan, 2012.