Xucan Chen
Research Subjects
Research Interests
Xucan Chen has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Hao Yu
Technical University of Munich, Mnchen, Germany
2 shared publications - Zheng Qin
Defense Innovation Institute, Academy of Military Sciences, Beijing, China
2 shared publications - Dianxi Shi
College of Computer Science and Technology, National University of Defense Technology, Changsha, China; Intelligent Game and Decision Lab (IGDL), Beijing, China
1 shared publication Jiayuan Zhuang
National University of Defense Technology
1 shared publication- Jiayuan Zhuang
Intelligent Game and Decision Lab (IGDL)
1 shared publication Tianqi Xu
National Innovation Institute of Defense Technology (NIIDT)
1 shared publication
Latest Publications
- 10 November 2023
- Published by Springer Nature
- p. 165-176
Reduced-Resolution Head for Object Detection
- Open Access
Task-Specific Context Decoupling for Object Detection
Classification and localization are two main sub-tasks in object detection. Nonetheless, these two tasks have inconsistent preferences for feature context, i.e., localization expects more boundary-aware features to accurately regress the bounding box, while more semantic context is preferred for object classification. Exsiting methods usually leverage disentangled heads to learn different feature context for each task. However, the heads are still applied on the same input features, which leads to an imperfect balance between classifcation and localization. In this work, we propose a novel Task-Specific COntext DEcoupling (TSCODE) head which further disentangles the feature encoding for two tasks. For classification, we generate spatially-coarse but semantically-strong feature encoding. For localization, we provide high-resolution feature map containing more edge information to better regress object boundaries. TSCODE is plug-and-play and can be easily incorperated into existing detection pipelines. Extensive experiments demonstrate that our method stably improves different detectors by over 1.0 AP with less computational cost. Our code and models will be publicly released.
Open AccessTask-Specific Context Decoupling for Object Detection
Classification and localization are two main sub-tasks in object detection. Nonetheless, these two tasks have inconsistent preferences for feature context, i.e., localization expects more boundary-aware features to accurately regress the bounding box, while more semantic context is preferred for object classification. Exsiting methods usually leverage disentangled heads to learn different feature context for each task. However, the heads are still applied on the same input features, which leads to an imperfect balance between classifcation and localization. In this work, we propose a novel Task-Specific COntext DEcoupling (TSCODE) head which further disentangles the feature encoding for two tasks. For classification, we generate spatially-coarse but semantically-strong feature encoding. For localization, we provide high-resolution feature map containing more edge information to better regress object boundaries. TSCODE is plug-and-play and can be easily incorperated into existing detection pipelines. Extensive experiments demonstrate that our method stably improves different detectors by over 1.0 AP with less computational cost. Our code and models will be publicly released.
- 22 January 2021
- Published by Springer Nature
- p. 340-353
Distributed Reinforcement Learning with States Feature Encoding and States Stacking in Continuous Action Space
- 22 January 2021
- Published by Springer Nature
- p. 340-353
Distributed Reinforcement Learning with States Feature Encoding and States Stacking in Continuous Action Space