Knowledge Commons of Institute of Automation,CAS
Retrieve the Visible Feature to Improve Thermal Pedestrian Detection Using Discrepancy Preserving Memory Network | |
Hu Yuxuan1,2![]() ![]() ![]() | |
2023-11 | |
会议名称 | 2023 IEEE International Conference on Image Processing |
会议日期 | 2023.10 |
会议地点 | Kuala Lumpur, Malaysia |
摘要 | We propose an approach for enhancing pedestrian detection in thermal infrared images using paired visible-thermal images in training. Recently, approaches that retrieve the corresponding visible features from thermal features using a key-value memory network have been proven effective for improving detection results. However, for memory networks storing thermal-visible features, random initialization and end-to-end training may not be ideal, as this can reduce the diversity of memory slots. Also, the retrieved visible features have different reliability as the overall similarities between key slots in the memory network and thermal features differ. These motivate us to propose a DIscrepancy Preserving (DIP) Memory that is updated manually to prevent convergence of key-value memory slots. We also evaluate the reliability of each retrieved visible feature and adjust the training protocol of the detection head. Experiment results on two visible-infrared pedestrian detection datasets demonstrate the superiority of our framework. |
关键词 | Thermal infrared pedestrian detection DIscrepancy Preserving (DIP) memory |
收录类别 | EI |
语种 | 英语 |
七大方向——子方向分类 | 目标检测、跟踪与识别 |
国重实验室规划方向分类 | 视觉信息处理 |
是否有论文关联数据集需要存交 | 否 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/56569 |
专题 | 多模态人工智能系统全国重点实验室_先进时空数据分析与学习 |
通讯作者 | Weng Lubin |
作者单位 | 1.State Key Laboratory of Multimodal Artificial Intelligence Systems, CASIA, Beijing, China 2.School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing, China 3.Shanghai Aerospace Electronic Technology Institute, Shanghai, China 4.Research Center of Aerospace Information, Institute of Automation, CAS, Beijing, China |
第一作者单位 | 中国科学院自动化研究所 |
推荐引用方式 GB/T 7714 | Hu Yuxuan,Zhang Ning,Weng Lubin. Retrieve the Visible Feature to Improve Thermal Pedestrian Detection Using Discrepancy Preserving Memory Network[C],2023. |
条目包含的文件 | ||||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
Retrieve_the_Visible(1702KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论