Unsupervised Pre-training for Person Re-identification

In this paper, we present a large scale unlabeled person re-identification (Re-ID) dataset "LUPerson" and make the first attempt of performing unsupervised pre-training for improving the generalization ability of the learned person Re-ID feature representation. This is to address the problem that all existing person Re-ID datasets are all of limited scale due to the costly effort required for data annotation. Previous research tries to leverage models pre-trained on ImageNet to mitigate the shortage of person Re-ID data but suffers from the large domain gap between ImageNet and person Re-ID data. LUPerson is an unlabeled dataset of 4M images of over 200K identities, which is 30X larger than the largest existing Re-ID dataset. It also covers a much diverse range of capturing environments (eg, camera settings, scenes, etc.). Based on this dataset, we systematically study the key factors for learning Re-ID features from two perspectives: data augmentation and contrastive loss. Unsupervised pre-training performed on this large-scale dataset effectively leads to a generic Re-ID feature that can benefit all existing person Re-ID methods. Using our pre-trained model in some basic frameworks, our methods achieve state-of-the-art results without bells and whistles on four widely used Re-ID datasets: CUHK03, Market1501, DukeMTMC, and MSMT17. Our results also show that the performance improvement is more significant on small-scale target datasets or under few-shot setting.

PDF Abstract CVPR 2021 PDF CVPR 2021 Abstract

Results from the Paper


Ranked #2 on Person Re-Identification on Market-1501 (using extra training data)

     Get a GitHub badge
Task Dataset Model Metric Name Metric Value Global Rank Uses Extra
Training Data
Result Benchmark
Person Re-Identification CUHK03 Unsupervised Pre-training (ResNet50+BDB) MAP 79.6 # 6
Rank-1 81.9 # 8
Person Re-Identification DukeMTMC-reID Unsupervised Pre-training (ResNet101+MGN) Rank-1 91.9 # 10
mAP 84.1 # 19
Person Re-Identification DukeMTMC-reID Unsupervised Pre-training (ResNet101+RK) Rank-1 93.99 # 4
mAP 92.77 # 3
Person Re-Identification Market-1501 Unsupervised Pre-training (ResNet101+RK) mAP 96.21 # 2
Person Re-Identification Market-1501 Unsupervised Pre-training (ResNet101+MGN) Rank-1 97 # 4
mAP 92 # 17
Person Re-Identification Market-1501-C LUPerson Rank-1 32.22 # 12
mAP 10.37 # 14
mINP 0.29 # 15
Person Re-Identification MSMT17 Unsupervised Pre-training (ResNet101+MGN) Rank-1 86.6 # 5
mAP 68.8 # 7

Methods


No methods listed for this paper. Add relevant methods here