Bolun Cai   蔡博仑 Email: caibolun@gmail.com LinkIn: https://www.linkedin.com/in/caibolun Github: https://github.com/caibolun/ Scholar: https://scholar.google.com/citations?user=wceVXfgAAAAJ

Biography

I am currently a engineer at Douyin Group. I received the M.Eng. and Ph.D. degrees from the South China University of Technology, China, in 2016 and 2019, respectively. My research interest mainly lies in computer vision, deep learning, image processing and large language model.

I am advised by Prof. Xiangmin Xu in SCUT and co-advised by Prof. Dacheng Tao in USYD. I am so lucky to work closely with Prof. Kui Jia in UM.

Education & Experience

• 2023 - Present    Douyin, ByteDance

Expert Researcher
• 2022 - 2023    Multiple Modal Understanding (MMU), Shopee
Expert Engineer
• 2019 - 2022    WeChatAI, Tencent
Senior Engineer
• 2016 - 2019    South China University of Technology
Ph.D., Communication and Information Engineering
• 2017.06 - 2017.09    WeChatAI, Tencent
Ph.D Researcher
• 2016.01 - 2016.09    University of Macau
Research Associate
• 2015.07 - 2015.09    DeepGlint Information Technologies Co. Ltd.
Computer Vision Engineer
• 2013 - 2016     South China University of Technology
M.E. , Electronics and Communications Engineering
• 2009 - 2013    South China University of Technology
B.E. , Integrated Circuit Design and Integration System

Theses

• Researches on Image Dehazing and Enhancement
  Bolun Cai
  A Dissertation of PhD, South China University of Technology (SCUT), 2019 [🏆 CIE Outstanding PhD]
  

  [CAJ] [PDF]

• Object Tracking Based on Biologically Inspired Model
  Bolun Cai
  A Dissertation of Master, South China University of Technology (SCUT), 2016
  

  [CAJ] [PDF]

• Biologically Inspired Model based Spatio-temporal Feature Extraction
  Bolun Cai
  A Dissertation of Bachelor, South China University of Technology (SCUT), 2013 [🏆 Excellent Graduated Thesis]
  

  [Paper]

Selected Publications [Google Scholar][dblp]

• Center Contrastive Loss for Metric Learning
  Bolun Cai, Pengfei Xiong, Shangxuan Tian.
  arXiv:2308.00458, 2023
  

  [Paper]

• MGAT: Multi-Granularity Attention Based Transformers for Multi-Modal Emotion Recognition
  Weiquan Fan, Xiaofen Xing, Bolun Cai, Xiangmin Xu.
  IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2023 [🏆 Top 3% Paper]
  

  [Paper] [Code]

• Unsupervised pre-training for detection transformers
  IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2022 [🏆 The First Prize of Guandong-CCF ]
  Zhigang Dai, Bolun Cai, Yugeng Lin, Junying Chen
  

  [Paper] [Code]

• UniMoCo: Unsupervised, Semi-Supervised and Fully-Supervised Visual Representation Learning
  IEEE International Conference on Systems, Man, and Cybernetics (SMC), 2022
  Zhigang Dai, Bolun Cai, Junying Chen
  

  [Paper] [Code]

• Unsupervised pre-training for detection transformers
  IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), 2022 [🏆 The Third Prize of Guandong-CCF]
  Lin Wang, Xiangmin Xu, Kailing Guo, Bolun Cai, Fang Liu
  

  [Paper]

• UP-DETR: Unsupervised pre-training for object detection with transformers
  IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2021
  Zhigang Dai, Bolun Cai, Yugeng Lin, Junying Chen
  

  [Paper] [Code]

• Multi-modality Hierarchical Recall based on GBDTs for Bipolar Disorder Classification
  Xiaofen Xing, Bolun Cai, Yinhu Zhao, Shuzhen Li, Zhiwei He, Weiquan Fan.
  Audio/Visual Emotion Challenge and Workshop, ACM Multimedia Conference , 2018 [🏆 The First Place]
  

  [Project] [Paper] [Slides] [Code]

• Perception Preserving Decolorization
  Bolun Cai, Xiangmin Xu, Xiaofen Xing.
  IEEE International Conference on Image Processing (ICIP), 2018
  

  [Project] [Paper] [Poster] [Code]

• A Joint Intrinsic-Extrinsic Prior Model for Retinex
  Bolun Cai, Xiangmin Xu, Kailing Guo, Kui Jia, Bin Hu, Dacheng Tao.
  IEEE International Conference on Computer Vision (ICCV), 2017
  

  [Project] [Paper] [Poster] [Code]

• DehazeNet: An End-to-End System for Single Image Haze Removal
  Bolun Cai, Xiangmin Xu, Kui Jia, Chunmei Qing, Dacheng Tao.
  IEEE Transactions on Image Processing (TIP), 2016 [🏆 Highly Cited & Hot Paper]
  

  [Project] [Paper] [Poster] [Code] [BReLU(Paddle)] [HardTanh(Torch)]

• BIT: Biologically Inspired Tracker
  Bolun Cai, Xiangmin Xu, Xiaofen Xing, Kui Jia, Jie Miao, Dacheng Tao.
  IEEE Transactions on Image Processing (TIP), 2016
  

  [Project] [Paper] [Poster] [Code]

• Edge/Structure Preserving Smoothing via Relativity-of-Gaussian
  Bolun Cai, Xiaofen Xing, Xiangmin Xu.
  IEEE International Conference on Image Processing (ICIP), 2017 [🏆 Best Paper Finalist]
  

  [Project] [Paper] [Slides] [Code]

• Real-time Video Dehazing based on Spatio-temporal MRF
  Bolun Cai, Xiangmin Xu, Dacheng Tao.
  Pacific-Rim Conference on Multimedia (PCM), 2016 [🏆 Best Student Paper]
  

  [Project] [Paper] [Slides] [Code]

• BIT: Bio-inspired Tracker
  Bolun Cai, Xiangmin Xu, Xiaofen Xing, Chunmei Qing.
  IEEE International Conference on Image Processing (ICIP), 2015
  

  [Project] [Paper] [Poster] [Code]

• Bio-inspired Model with Dual Visual Pathways for Human Action Recognition
  Bolun Cai, Xiangmin Xu, Chunmei Qing.
  International Symposium on Communication Systems, Networks & Digital Signal Processing (CSNDSP), 2014
  

  [Paper] [Slides]

• A Coarse-to-Fine Approach for Motion Pattern Discovery
  Bolun Cai, Zhifeng Luo, Kerui Li.
  International Conference on Cyber-Enabled Distributed Computing and Knowledge Discovery (CyberC), 2012
  

  [Paper] [Slides]

• FReLU: Flexible Rectified Linear Units for Improving Convolutional Neural Networks
  Suo Qiu, Xiangmin Xu, Bolun Cai.
  IEEE International Conference on Pattern Recognition (ICPR), 2018
  

  [Paper] [Code]

• Multi-invariance appearance model for object tracking
  Guicong Xu, Xiangmin Xu, Xiaofen Xing, Bolun Cai, Chunmei Qing.
  IEEE International Conference on Digital Signal Processing (DSP), 2015
  

  [Paper] [Report]