About me

Hi! I am currently a postdoctoral research fellow at NExT++, School of Computing, National University of Singapore with Prof. Tat-Seng Chua.

2017 - 2021 I received Ph.D. from the ReLER Lab, University of Technology Sydney (UTS) , under the supervision of Prof. Yi Yang and Dr. Liang Zheng.

2012 - 2016 I received my Bachelor’s degree from Fudan University in 2016, under the supervision of Prof. Xiangyang Xue.

2018, 2020 I was fortunate to spend my summer working with Xiaodong Yang (Nvida), Zhiding Yu (Nvidia), Jan Kautz (Nvidia), Minyue Jiang (Baidu) and Xiao Tan (Baidu).

Open-source projects can be found at my [Github], and publications can be found at SCI, SCOPUS, or Google Scholar.

More details can be found in my [CV].

If you are a NUS Undergraduate / Master student who is interested in doing research/project with me, please contact me via email with your CV. My current email address is zdzheng AT nus.edu.sg .


Research Statement

My work focuses on image retrieval and image generation.

Big data is the primary part of training data-driven models. There remain three scientific questions.

  • Data Generation: How to obtain more data? Due to the annotation costs and privacy concerns, we usually could not access the large-scale data easily.

  • Prior Knowledge: Does more data mean a better model? Deep learning also demands a deep understanding of data (robustness and explainability).

  • Efficiency: How to train on million-scale data? What data matters most? Efficient training and inference is needed.



News

  • We are holding the special session at ICME 2022 on Beyond Accuracy: Responsible, Responsive, and Robust Multimedia Retrieval. [Call for papers]
  • My Ph.D. thesis was on the Chancellor's List.
  • I was awarded 2021 IEEE Circuits and Systems Society Outstanding Young Author Award. Thanks a lot for supports and understanding from my supervisors and friends. [link]
  • One paper on Uncertainty is accepted by IJCV 2021. [PDF] [code]
  • We have released a new Drone-view Geo-localization Dataset, ACM Multimedia 2020. [PDF] [Dataset] [Video] [Slide]
  • People live in a 3D world. Why not conduct representation learning in the 3D space? [arXiv] [code]
  • We have achieved the 1st place in AICity Challenge Vehicle Re-id Track, CVPR 2020. [code]
  • Two papers to appear at IJCAI 2020. [PDF1][code1][PDF2] [code2]
  • One paper to appear at CVPR 2019 as oral presentation. [PDF][3-min video] [code]

  • Others

  • Shanghai is my hometown, and it is a lovely place to have a sightseeing tour.
  • I was a poster maker when I studied at Fudan University. You may check out them at Here.
  • I sometimes write Chinese blogs and share insights at Zhihu.