Multi-View Learning for Vehicle Re-Identification | |
Weipeng Lin1; Yidong Li1; Xiaoliang Yang1; Peixi Peng2; Junliang Xing2 | |
2019 | |
会议日期 | July 8-12, 2019 |
会议地点 | Shanghai, China |
英文摘要 | Vehicle re-identification (ReID) aims to identify a target vehicle in different cameras with non-overlapping views, and it plays an important role when the car licence plate recognition is unavailable or unreliable. Compared with face recognition and person ReID tasks, it is difficult to train an effective vehicle ReID model due to two reasons: the different views greatly affect the visual appearance of a vehicle, and different vehicles may exhibit fairly similar visual appearance when their images are captured from one unified single view. To handle this training difficulty, we introduce several latent groups to represent multiple views. Then, the vehicle ReID problem is modeled as two sub tasks including matching vehicles in a same view and across different views. A fine-grain ranking loss and a relative coarse-grain ranking loss are proposed to each task respectively. Extensive experimental analyses and evaluations on two benchmarks demonstrate the proposed method can achieve state-of-the-art performance. |
内容类型 | 会议论文 |
源URL | [http://ir.ia.ac.cn/handle/173211/26157] |
专题 | 中国科学院自动化研究所 |
通讯作者 | Xiaoliang Yang |
作者单位 | 1.Beijing Jiaotong University 2.Chinese Academy of Sciences, Beijing |
推荐引用方式 GB/T 7714 | Weipeng Lin,Yidong Li,Xiaoliang Yang,et al. Multi-View Learning for Vehicle Re-Identification[C]. 见:. Shanghai, China. July 8-12, 2019. |
个性服务 |
查看访问统计 |
相关权益政策 |
暂无数据 |
收藏/分享 |
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