The Skyline of Hang

Let's explore the world of AI.

About Me

I am an assistant professor working in the Department of Computer Science and Technology at Tsinghua University, working with Prof. Bo Zhang and Prof. Jun Zhu. My research interests lie in the development of computer vision and machine learning algorithms for solving scientific and engineering problems. My current work involves both the foundations of interpretable machine learning and the applications of image/video analysis. Before joining Tsinghua, I received my Ph. D. degree from Shanghai Jiaotong University and worked as a visiting scholar at Carnegie Mellon University 

Dr.  Hang   Su (苏 航)

Assist. Prof.  @ THU


July 2018
  • Our paper "Essay-Anchor Attentive Multi-Modal Bilinear Pooling for Textbook Question Answering" wins ICME2018 Platinum Best Paper Award! 
  • One paper on Video to Video translation is accepted by ACM Multimedia 2018.

June 2018

May 2018

Apr. 2018
  • One paper on Chinese Character Generation with disentangled deep generative model is accepted by IJCAI2018.  
Mar. 2018
  • One paper on TextBook Question Answering by symbolic-neural reasoning is accepted by ICME2018 as oral. 
  • One proposal on "Deep Reinforcement Learning with applications on Video Games " was funded by Tencent AI Lab Rhino-Bird Joint Research Program.
Feb. 2018
  • One paper on interpreting DNNs by identifying the critical data routing paths is accepted by CVPR2018 as spotlight. 

  • One paper on joint reasoning over images and symbolic texts for TQA is accepted by CVPR2018 as spotlight. 

  • One paper on boosting the performance of adversarial attacks with momentum is accepted by CVPR2018 as spotlight.

  • One paper on understanding crowd behaviors by imitating the decision-making process  is presented in AAAI2018. 

  • One paper on  person re-identification based on  graph correspondence transfer is presented in AAAI2018.

Jan. 2018
  • One proposal on " Explainable Machine Learning" was funded by  MSRA Joint Research. 
  • One proposal on " Interpretable Deep Learning and Visual Relation Understanding " was funded by Intel Labs China.
  • One paper on  facial  sketch retrieval by generating photorealistic faces with conditional GAN is accepted by ICASSP2018.