About me

I am a third-year undergradutate in School of EECS at Peking University, majoring in Computer Science advised by Prof. Zhanxing Zhu. I have been a research intern at Megvii Inc.(Face++) mentored by Dr. Xiangyu Zhang since March, 2018. Currently visiting Berkeley!

Research topics

  • Computer vision and deep learning, especially visual recognitions, maniuplating images.
  • Efficient learning and inference. Speed up the training of NNs. AutoML, NAS for light-weight NNs.
  • Dynamic networks.

News


[July, 2019] One paper accepted by ICCV19. Comming soon.

[April, 2019] One paper accepted by ICML19. Codes and pdf comming soon.

[April. 2019] We published a new general detection dataset Objects365, which is designed to spur object detection research with a focus on diverse objects in the Wild.

Publications


You Only Propagate Once: Accelerating Adversarial Training via Maximal Principle. Dinghuai Zhang*, Tianyuan Zhang*,Yiping Lu*, Zhanxing Zhu, Bin Dong. (*equal contribution) [code]

Interpreting Adversarially Trained Convolutional Neural Networks. Tianyuan Zhang, Zhanxing Zhu. [code] International Conference on Machine Learning (ICML), 2019.