TripleA: An Unsupervised Domain Adaptation Framework for Nighttime VRU Detection
发布时间：2025-06-01
点击次数：
- DOI码：
- 10.1109/tits.2025.3548804
- 所属单位：
- Wuhan Univ, State Key Lab Informat Engn Surveying Mapping & Re, Wuhan 430079, Peoples R China
- 发表刊物：
- IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
- 项目来源：
- This work was supported in part by Shanxi Provincial Science and Technology Major Special Project un
- 关键字：
- Lighting,Noise,Degradation,Noise reduction,Training,Image enhancement,Annotations,Detectors,Atmospheric modeling,Roads,Vulnerable road user detection,unsupervised domain adaptation,image enhancement,low-light,denoising
- 摘要：
- Detecting vulnerable road users (VRUs) at night presents significant challenges. Numerous methods rely heavily on annotations, yet the low visibility of nighttime images poses difficulties for labeling. To obviate the need for nighttime annotations, unsupervised domain adaptation manifests as a viable solution. However, existing approaches primarily focus on semantic-level domain gaps, often overlooking pixel-level discrepancies caused by inherent degradations in the nighttime domain. These degradations can impair machine vision and limit detection performance. In this paper, we propose Triple
- 合写作者：
- Wang, Jiaming,Wang, Yu,Ding, Yulin,Cheng, Gui
- 第一作者：
- Wang, Yuankun
- 论文类型：
- 期刊论文
- 通讯作者：
- Shao, Zhenfeng
- 文献类型：
- Article
- 卷号：
- 26
- 期号：
- 6
- 页面范围：
- 8320-8336
- ISSN号：
- 1524-9050
- 是否译文：
- 否
- 发表时间：
- 2025-06-01