Journal
2021 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2021)
Volume -, Issue -, Pages 1-10Publisher
IEEE COMPUTER SOC
DOI: 10.1109/WACV48630.2021.00005
Keywords
-
Ask authors/readers for more resources
A neural network with an asymmetric structure is proposed to solve the problem of local minima caused by label noise in unsupervised Re-ID. By using asymmetric branches and cross-branch supervision, feature diversity and weight diversity between teacher and student networks are enhanced.
The objective of unsupervised person re-identification (Re-ID) is to learn discriminative features without labor-intensive identity annotations. State-of-the-art unsupervised Re-ID methods assign pseudo labels to unlabeled images in the target domain and learn from these noisy pseudo labels. Recently introduced Mean Teacher Model is a promising way to mitigate the label noise. However, during the training, self-ensembled teacher-student networks quickly converge to a consensus which leads to a local minimum. We explore the possibility of using an asymmetric structure inside neural network to address this problem. First, asymmetric branches are proposed to extract features in different manners, which enhances the feature diversity in appearance signatures. Then, our proposed cross-branch supervision allows one branch to get supervision from the other branch, which transfers distinct knowledge and enhances the weight diversity between teacher and student networks. Extensive experiments show that our proposed method can significantly surpass the performance of previous work on both unsupervised domain adaptation and fully unsupervised Re-ID tasks.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available