Dongliang Chang

Dongliang Chang 常东良

Ph.D. Candidate
Pattern Recognition and Intelligent System Laboratory (PRIS lab)
School of Artificial Intelligence
Beijing University of Posts and Telecommunications (BUPT)

Email: changdongliang@bupt.edu.cn
WeChat: Dongliang Chang
[CV] [Google Scholar] [GitHub] [Team GitHub] [ArXiv]


My primary research interests are Fine-Grained Image Analysis (FGIA) and Domain adaptation (DA).
In particular, I an interested in developing algorithms that can understand what people see and contribute to a better life.

I am advised by Prof. Zhanyu Ma and Prof. Yi-Zhe Song.

Keywords: Fine-Grained Image Analysis, Domain Adaptation, Transfer Learning, Open Set


News

April 2020

Our one paper is accepted at TIP 2020!

Mar 2020

Released my paper "Mind the Gap: Enlarging the Domain Gap in Open Set Domain Adaptation " on ArXiv [paper] [code]!

Mar 2020

Our one paper is accepted at ICME 2020!

Feb 2020

Our one paper is accepted at TIP 2020! See CSIG 成果速览 for media coverage.


Education

Beijing University of Posts and Telecommunications (BUPT) (From 2019.09)

  • Ph.D. candidate of PRIS lab, BUPT
  • Advisor: Prof. Zhanyu Ma

  • Beijing University of Posts and Telecommunications (BUPT) (2017.04-2019.09)
  • Visiting Student
  • Advisor: Prof. Zhanyu Ma

  • Lanzhou University of Technology (LUT) (2016.09-2019.06)
  • M.E. in IoT Engineering
  • Advisor: Prof. Xiaoxu Li

  • Zhoukou Normal University (ZKNU) (2012.09 ~ 2016.06)
  • B.E. in Network engineering
  • Advisor: Prof. Qi Wang

  • Research Project

    基于域适应的细粒度图像分类方法研究,主持,北京邮电大学博士生创新基金 (BUPT Excellent Ph.D. Students Foundation),项目批准号:XX, 2020.1-2022.1. [8/100]

    Honors

    2019 年首届北京邮电大学优秀博士生后备计划一等奖学金 [32/342]

    Professional Services

    Reviewer:
    IEEE Transactions on Vehicular Technology 2019, 2020
    NEUROCOMPUTING 2020
    ICPR 2020

    Publications

    International Journal

    • OSLNet: Deep Small-Sample Classification with an Orthogonal Softmax Layer
      Xiaoxu Li, Dongliang Chang, Zhanyu Ma, Zheng-Hua Tan, Jing-Hao Xue, Jie Cao, Jingyi Yu, and Jun Guo
      IEEE Transactions on Image Processing 2020
      [paper] [code]

    • The Devil is in the Channels: Mutual-Channel Loss for Fine-Grained Image Classification
      Dongliang Chang, Yifeng Ding, Jiyang Xie, Ayan Kumar Bhunia, Xiaoxu Li, Zhanyu Ma*, Ming Wu, Jun Guo, and Yi-Zhe Song
      IEEE Transactions on Image Processing 2020
      [paper] [code]

    • Large-margin Regularized Softmax Cross-Entropy Loss
      Xiaoxu Li*, Dongliang Chang, Tao Tian, and Jie Cao
      IEEE Access 2019
      [paper]

    • Fine-Grained Vehicle Classification with Channel Max Pooling Modified CNNs
      Zhanyu Ma*, Dongliang Chang, Jiyang Xie, Yifeng Ding, Shaoguo Wen, Xiao-Xu Li, Zhongwei Si, and Jun Guo
      IEEE Transactions on Vehicular Technology 2019
      [paper]

    • Dual Cross-Entropy Loss for Small-Sample Fine-Grained Vehicle Classification
      Xiaoxu Li*, Liyun Yu, Dongliang Chang, Zhanyu Ma*, and Jie Cao
      IEEE Transactions on Vehicular Technology 2019
      [paper]

    • Prediction of short-term PV power output and uncertainty analysis
      Luyao Liu, Yi Zhao, Dongliang Chang, Jiyang Xie, Zhanyu Ma*, Qie Sun*, Hongyi Yin*, and Ronald Wennersten
      Applied Energy 2018
      [paper]

    International Conference

    • IU-Module: Intersection and Union Module for Fine-Grained Visual Classification
      Yixiao Zheng, Dongliang Chang, Jiyang Xie, and Zhanyu Ma*
      IEEE International Conference on Multimedia and Expo (ICME), 2020

    • FICAL: Focal Inter-Class Angular Loss for Image Classification
      Xinran Wei, Dongliang Chang, Jiyang Xie, Yixiao Zheng, Chen Gong, Chuang Zhang, and Zhanyu Ma
      IEEE International Conference on Visual Communications and Image Processing (VCIP), 2019
      [paper]

    • Channel Max Pooling for Image Classification
      Lu Cheng, Dongliang Chang*, Jiyang Xie, Rongliang Ma, Chunsheng Wu, and Zhanyu Ma
      International Conference on Intelligence Science and Big Data Engineering (IScIDE), 2019
      [paper]

    • Dynamic Attention Loss for Small-sample Image Classification
      Jie Cao, Yinping Qiu, Dongliang Chang, Xiaoxu Li*, and Zhanyu Ma*
      The 11th Annual Conference Organized by Asia-Pacific Signal and Information Processing Association (APSIPA), 2019
      [paper]

    • Mixed Attention Mechanism for Small-Sample Fine-grained Image Classification
      Xiaoxu Li, Jijie Wu, Dongliang Chang, Zhanyu Ma*, and Jie Cao*
      The 11th Annual Conference Organized by Asia-Pacific Signal and Information Processing Association (APSIPA), 2019
      [paper]

    • Small-Sample Image Classification Method of Combining Prototype and Margin Learning
      Xiaoxu Li, Liyun Yu, Dongliang Chang, Zhanyu Ma*, and Jie Cao*
      The 11th Annual Conference Organized by Asia-Pacific Signal and Information Processing Association (APSIPA), 2019
      [paper]

    • SSE: A new selective initialization strategy for Snapshot Ensembling
      Dongliang Chang, Xiaoxu Li*, Jiyang Xie, Zhanyu Ma, Jun Guo, and Jie Cao
      IEEE International Conference on Cloud Computing and Intelligence Systems (CCIS), 2018
      [paper]

    • Softmax Cross Entropy Loss with Unbiased Decision Boundary for Image Classification
      Jie Cao*, Zhe Su, Liyun Yu, Dongliang Chang, Xiaoxu Li , and Zhanyu Ma
      Chinese Automation Congress(CAC), 2018
      [paper]

    Arxiv

    • Mind the Gap: Enlarging the Domain Gap in Open Set Domain Adaptation
      Dongliang Chang, Aneeshan Sain, Zhanyu Ma, Yi-Zhe Song, Jun Guo
      ArXiv 2020
      [paper] [code]

    • Dual-attention Guided Dropblock Module for Weakly Supervised Object Localization
      Junhui Yin, Siqing Zhang, Dongliang Chang, Zhanyu Ma, Jun Guo
      ArXiv 2020
      [paper]

    • Fine-Grained Visual Classification via Progressive Multi-Granularity Training of Jigsaw Patches
      Ruoyi Du, Dongliang Chang, Ayan Kumar Bhunia, Jiyang Xie, Yi-Zhe Song, Zhanyu Ma, Jun Guo
      ArXiv 2020
      [paper] [code]

    • Channel Attention with Embedding Gaussian Process: A Probabilistic Methodology
      Jiyang Xie, Dongliang Chang, Zhanyu Ma, Guoqiang Zhang, Jun Guoo
      ArXiv 2020
      [paper]

    • Weakly Supervised Attention Pyramid Convolutional Neural Network for Fine-Grained Visual Classification
      Yifeng Ding, Shaoguo Wen, Jiyang Xie, Dongliang Chang, Zhanyu Ma, Zhongwei Si, Haibin Lingn
      ArXiv 2020
      [paper] [code]

    • Competing Ratio Loss for Discriminative Multi-class Image Classification
      Ke Zhang, Xinsheng Wang, Yurong Guo, Dongliang Chang, Zhenbing Zhao, Zhanyu Ma, Tony X.Han
      ArXiv 2019
      [paper]

    * denotes corresponding author.

    Dataset

    Overview
    EEG-Database involves three parts: 1) individual behavior data; 2) individual EEG data; 3) image-sketch-text data. Individual behavior data and individual EEG data are collected from the experiments on 24 subjects. Image-sketch-text data contains 1120 image-sketch-text pairs, which are split into 4 parts randomly. More details can be found in the README of EEG-Database.

    Dataset
    Download
    All data can be downloaded [here]. (passwd: nd0w)