Bingchen Zhao

This site is deprecated, please visit my new website I am a third year undergrad student major in Computer Science at Tongji University, where I worked closely with Prof. Yin Wang. I started my research internship at Megvii Research Nanjing on Feb 2020, where I worked closely with Dr. Xin Jin.

I open source as much as I can, you can find the code of my work on Github.

Email / GitHub / Google Scholar

My research interests include but not limited to

Self-Supervised Learning
Semi-Supervised Learning
Domain Adaptation

News

         ✩ [Jul. 2020] Our team ranked 2nd in the VIPriors Image Classification challenge ECCV2020.
         ✩ [Jun. 2020] Our team wins the 1st place of iWildCam2020 FGVC7 CVPR2020.
         ✩ [Apr. 2020] I have one paper accepted by APWeb2020 conference.
         ✩ [Apr. 2020] Our team ranked 8th at the 1st Agriculture-Vision Challenge CVPR2020.
         ✩ [Apr. 2020] I have one paper accepted by CVPR2020 workshop proceedings.

Publications

(*) indicates equal contribution.

Characterizing Robotic and Organic Query in SPARQL Search Sessions

Xinyue Zhang, Meng Wang, Bingchen Zhao, Ruyang Liu, Jingyuan Zhang, Han Yang
The 4th APWeb-WAIM International Joint Conference on Web and Big Data, 2020

We proposed a method for detecting and separating robotic SPARQL queries based on session-level query features.

Reducing the feature divergence of RGB and near-infrared images using Switchable Normalization

Siwei yang*, Shaozuo Yu*, Bingchen Zhao*, Yin Wang
Proceedings of IEEE CVPR Workshops, 2020
code / slides /

We consider the feature divergence between RGB and NIR channels in agriculture data, switchable norm is used to improve the performance.

Preprints

These paper are still working in progress.

Context Encoding for Video Retrieval with Contrastive Learning

Jie Shao*, Xin Wen*, Bingchen Zhao, Changhu Wang, Xiangyang Xue
, 2020
arxiv /

We consider the feature divergence between RGB and NIR channels in agriculture data, switchable norm is used to improve the performance.

Distilling Visual Priors from Self-Supervised Learning

Bingchen Zhao, Xin Wen
, 2020
arxiv / pdf / code / slides /

We present a novel two-phase pipeline that leverages self-supervised learning and knowledge distillation to improve the generalization ability of CNN models for image classification under the data-deficient setting.

Misc

Design and source code from Jon Barron's website