Unsupervised Cross-Dataset Transfer Learning for Person Re-identification
文献类型：会议论文
| 作者 | Peng PX(彭佩玺)1 ; Tao Xiang3; Wang YW(王耀威)4; Massimiliano Pontil2; Huang TJ(黄铁军)1; Tian YH(田永鸿)1
|
| 出版日期 | 2016 |
| 会议日期 | 2016 .07 |
| 会议地点 | Las Vegas, NV, USA |
| 英文摘要 | Most existing person re-identification (Re-ID) approaches follow a supervised learning framework, in which a large number of labelled matching pairs are required for training. This severely limits their scalability in realworld applications. To overcome this limitation, we develop a novel cross-dataset transfer learning approach to learn a discriminative representation. It is unsupervised in the sense that the target dataset is completely unlabelled. Specifically, we present an multi-task dictionary learning method which is able to learn a dataset-shared but targetdata-biased representation. Experimental results on five benchmark datasets demonstrate that the method significantly outperforms the state-of-the-art. |
| 源URL | [http://ir.ia.ac.cn/handle/173211/20219]  |
| 专题 | 自动化研究所_模式识别国家重点实验室_视频内容安全团队 |
| 作者单位 | 1.北京大学 2.University College London 3.Queen Mary, Univ. of London, 4.北京理工大学 |
| 推荐引用方式 GB/T 7714 | Peng PX,Tao Xiang,Wang YW,et al. Unsupervised Cross-Dataset Transfer Learning for Person Re-identification[C]. 见:. Las Vegas, NV, USA. 2016 .07. |
入库方式： OAI收割
来源：自动化研究所
浏览0
下载0
收藏0
其他版本
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。