| 引用本文: | 张艳,相旭,唐俊,等.跨模态行人重识别的对称网络算法.[J].国防科技大学学报,2022,44(1):122-128.[点击复制] |
| ZHANG Yan,XIANG Xu,TANG Jun,et al.Cross-modality person re-identification algorithm using symmetric network[J].Journal of National University of Defense Technology,2022,44(1):122-128[点击复制] |
|
| |
|
|
| 本文已被：浏览 5767次 下载 4214次 |
| 跨模态行人重识别的对称网络算法 |
| 张艳,相旭,唐俊,王年,屈磊 |
|
(安徽大学 电子信息工程学院, 安徽 合肥 230601)
|
| 摘要: |
| 针对模态间差异,提出基于对称网络的跨模态行人重识别算法,该网络将基于概率分布的模态混淆与对抗学习结合,通过对称网络产生模态不变特征,从而达到模态混淆的目的；针对外观差异和模态内差异,使用不同隐藏层的网络卷积特征构造混合三元损失,提高网络的特征表征能力。RegDB和SYSU-MM01数据集上的大量实验结果表明了该方法的有效性。 |
| 关键词: 跨模态 行人重识别 对称网络 对抗学习 混合三元损失 |
| DOI：10.11887/j.cn.202201018 |
| 投稿日期：2020-07-22 |
| 基金项目:国家自然科学基金资助项目(61772032,61871411)；国家重点研发计划资助项目(2018YFC0807302) |
|
| Cross-modality person re-identification algorithm using symmetric network |
| ZHANG Yan, XIANG Xu, TANG Jun, WANG Nian, QU Lei |
|
(School of Electronic and Information Engineering, Anhui University, Hefei 230601, China)
|
| Abstract: |
| For the difference between modalities, a cross-modality person re-identification algorithm which based on symmetric network was proposed. The network combined the modal confusion based on probability distribution with adversarial learning, and generated modal-invariant features through symmetric network to achieve modal confusion. To deal with appearance differences and intra-modality differences, the network constructed a mixed-triplet loss using convolution features of different hidden layers, which can improve the characterization capability of the network. Numerous experimental results on the RegDB and SYSU-MM01 datasets demonstrate the effectiveness of the method. |
| Keywords: cross-modality person re-identification symmetric network adversarial learning mixed-triplet loss |
|
|
|
|
|
