Haiyang Mei


Ph.D Student
Dalian University of Technology

Email: mhy666@mail.dlut.edu.cn


I am currently a second-year Ph.D. student in the School of Computer Science and Technology, Dalian University of Technology, supervised by Prof Xiaopeng Wei and Prof Xin Yang, and also received academic guidance from Rynson W.H. Lau and Bo Dong. Before that, I received the B.Eng. degree from the School of Control Science and Engineering, Dalian University of Technology. in 2017.

Research Interests

My primary research interest is in designing effective visual understanding models for the vision systems. This work can help AI agents in scene-level understanding, reasoning, and decision making.

My recent endeavor is on scene confusing-discovery, which aims to mine the confusing/special yet meaningful object/region in the scene. In particular, the objects I focus on include glass, mirror, camouflaged object, and salient object, which are very common in daily life scenes but can confuse the vision systems due to their inherently special properties. Therefore, detecting and segmenting such objects from the scene plays an essential role in accurate scene understanding and can benefit a wide range of computer vision, graphics, and multimedia applications, including image classification, visual tracking, content-aware image editing, medical image diagnosis, and robotic navigation. However, this task has not been fully explored and remains an unsolved and challenging problem. Both glass and mirror do not have their own visual appearances but only transmit/reflect the appearances of their surroundings, making them fundamentally different from other common objects that have been addressed well by the state-of-the-art segmentation methods. The camouflaged/salient object is the object that is ``seamlessly'' embedded in their surroundings or most attention-grabbing and could be easily cluttered by the background in the complex scenes. I am working toward exploring useful cues and effective methods for accurate segmentation.

Besides, my early work was on image super-resolution, which is to reconstruct the high-quality, visually satisfactory high-resolution image from the input low-resolution one and is the cornerstone of providing more detailed information for scene analysis and understanding.


Publications  [Google Scholar]

Depth-Aware Mirror Segmentation
Haiyang Mei, Bo Dong, Wen Dong, Pieter Peers, Xin Yang, Qiang Zhang, Xiaopeng Wei
The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021 (Oral)
[PDF|Project Page]
Camouflaged Object Segmentation with Distraction Mining
Haiyang Mei, Ge-Peng Ji, Ziqi Wei, Xin Yang, Xiaopeng Wei, Deng-Ping Fan
The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021
[PDF|Project Page]
Exploring Dense Context for Salient Object Detection
Haiyang Mei, Yuanyuan Liu, Ziqi Wei, Dongsheng Zhou, Xiaopeng Wei, Qiang Zhang, Xin Yang
IEEE Transactions on Circuits and Systems for Video Technology (TCSVT) 2021
[PDF|Project Page]
A Two-Stage Attentive Network for Single Image Super-Resolution
Jiqing Zhang, Chengjiang Long, Yuxin Wang, Haiyin Piao, Haiyang Mei, Xin Yang, Baocai Yin
IEEE Transactions on Circuits and Systems for Video Technology (TCSVT) 2021
[PDF|Project Page]

Don't Hit Me! Glass Detection in Real-World Scenes
Haiyang Mei, Xin Yang, Yang Wang, Yuanyuan Liu, Shengfeng He,
Qiang Zhang, Xiaopeng Wei, Rynson W.H. Lau
The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2020
[PDF|Project Page]
Multi-Context And Enhanced Reconstruction Network For Single Image Super Resolution
Jiqing Zhang, Chengjiang Long, Yuxin Wang, Xin Yang, Haiyang Mei, and Baocai Yin
IEEE International Conference on Multimedia and Expo (ICME) 2020
[PDF|Project Page]

Where Is My Mirror?
Xin Yang*, Haiyang Mei*, Ke Xu, Xiaopeng Wei, Baocai Yin, and Rynson Lau. (* joint first authors)
IEEE International Conference on Computer Vision (ICCV) 2019
[PDF|Project Page]

DRFN: Deep Recurrent Fusion Network for Single Image Super-Resolution with Large Factors
Xin Yang, Haiyang Mei, Jiqing Zhang, Ke Xu, Baocai Yin, Qiang Zhang, and Xiaopeng Wei.
IEEE Transactions on Multimedia (TMM) 2018
[PDF| Project Page]


1.   一种利用回归树场的图像超分辨率放大方法.   杨鑫, 梅海洋, 许可, 魏小鹏, 尹宝才.   201710859709.0

Website visit statistics