Unsupervised Long-Term Person Re-Identification with Clothes Change (2202.03087v3)
Abstract: We investigate unsupervised person re-identification (Re-ID) with clothes change, a new challenging problem with more practical usability and scalability to real-world deployment. Most existing re-id methods artificially assume the clothes of every single person to be stationary across space and time. This condition is mostly valid for short-term re-id scenarios since an average person would often change the clothes even within a single day. To alleviate this assumption, several recent works have introduced the clothes change facet to re-id, with a focus on supervised learning person identity discriminative representation with invariance to clothes changes. Taking a step further towards this long-term re-id direction, we further eliminate the requirement of person identity labels, as they are significantly more expensive and more tedious to annotate in comparison to short-term person re-id datasets. Compared to conventional unsupervised short-term re-id, this new problem is drastically more challenging as different people may have similar clothes whilst the same person can wear multiple suites of clothes over different locations and times with very distinct appearance. To overcome such obstacles, we introduce a novel Curriculum Person Clustering (CPC) method that can adaptively regulate the unsupervised clustering criterion according to the clustering confidence. Experiments on three long-term person re-id datasets show that our CPC outperforms SOTA unsupervised re-id methods and even closely matches the supervised re-id models.
- “Unsupervised tracklet person re-identification,” TPAMI, 2019.
- “Learning shape representations for clothing variations in person re-identification,” arXiv preprint arXiv:2003.07340, 2020.
- “Person re-identification by video ranking,” in ECCV. Springer, 2014.
- “Person re-identification by contour sketch under moderate clothing change,” TPAMI, 2019.
- “Deepchange: A long-term person re-identification benchmark,” arXiv preprint arXiv:2105.14685, 2021.
- “Unsupervised person re-identification via softened similarity learning,” in CVPR, 2020.
- “Cluster-guided asymmetric contrastive learning for unsupervised person re-identification,” arXiv preprint arXiv:2106.07846, 2021.
- “Fine-grained shape-appearance mutual learning for cloth-changing person re-identification,” in CVPR, 2021.
- “Automatic curriculum learning for deep rl: A short survey,” arXiv preprint arXiv:2003.04664, 2020.
- “On the power of curriculum learning in training deep networks,” in ICML, 2019.
- “Deep residual learning for image recognition,” in CVPR, 2016.
- “A density-based algorithm for discovering clusters in large spatial databases with noise,” in SIGKDD, 1996.
- “Self-paced contrastive learning with hybrid memory for domain adaptive object re-id,” in NeurIPS, 2020.
- “Cluster contrast for unsupervised person re-identification,” arXiv preprint arXiv:2103.11568, 2021.
- “Mobilenetv2: Inverted residuals and linear bottlenecks,” in CVPR, 2018.
- “Learning generalisable omni-scale representations for person re-identification,” TPAMI, 2021.
- “Densely connected convolutional networks,” in CVPR, 2017.
- “Beyond scalar neuron: Adopting vector-neuron capsules for long-term person re-identification,” TCSVT, 2019.
- “Training data-efficient image transformers & distillation through attention,” in ICML, 2021.
- “A strong baseline and batch normalization neck for deep person re-identification,” TMM, 2020.
- “An image is worth 16x16 words: Transformers for image recognition at scale,” arXiv preprint arXiv:2010.11929, 2020.
- “Rethinking the inception architecture for computer vision,” in CVPR, 2016.
- “Cluster-guided asymmetric contrastive learning for unsupervised person re-identification,” IEEE Transactions on Image Processing, vol. 31, pp. 3606–3617, 2022.
- Mingkun Li (6 papers)
- Shupeng Cheng (4 papers)
- Peng Xu (357 papers)
- Xiatian Zhu (139 papers)
- Chun-Guang Li (22 papers)
- Jun Guo (130 papers)