Transfer and share: semi-supervised learning from long-tailed data.
Saved in:
| Title: | Transfer and share: semi-supervised learning from long-tailed data. |
|---|---|
| Authors: | Wei, Tong1 (AUTHOR) weit@seu.edu.cn, Liu, Qian-Yu2 (AUTHOR), Shi, Jiang-Xin2 (AUTHOR), Tu, Wei-Wei3 (AUTHOR), Guo, Lan-Zhe2 (AUTHOR) guolz@lamda.nju.edu.cn |
| Source: | Machine Learning. Apr2024, Vol. 113 Issue 4, p1725-1742. 18p. |
| Subjects: | Supervised learning, Secure Sockets Layer (Computer network protocol), Logits, Sharing |
| Abstract: | Long-Tailed Semi-Supervised Learning (LTSSL) aims to learn from class-imbalanced data where only a few samples are annotated. Existing solutions typically require substantial cost to solve complex optimization problems, or class-balanced undersampling which can result in information loss. In this paper, we present the TRAS (TRAnsfer and Share) to effectively utilize long-tailed semi-supervised data. TRAS transforms the imbalanced pseudo-label distribution of a traditional SSL model via a delicate function to enhance the supervisory signals for minority classes. It then transfers the distribution to a target model such that the minority class will receive significant attention. Interestingly, TRAS shows that more balanced pseudo-label distribution can substantially benefit minority-class training, instead of seeking to generate accurate pseudo-labels as in previous works. To simplify the approach, TRAS merges the training of the traditional SSL model and the target model into a single procedure by sharing the feature extractor, where both classifiers help improve the representation learning. According to extensive experiments, TRAS delivers much higher accuracy than state-of-the-art methods in the entire set of classes as well as minority classes. Code for TRAS is available at https://github.com/Stomach-ache/TRAS. [ABSTRACT FROM AUTHOR] |
| Copyright of Machine Learning is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
| Abstract: | Long-Tailed Semi-Supervised Learning (LTSSL) aims to learn from class-imbalanced data where only a few samples are annotated. Existing solutions typically require substantial cost to solve complex optimization problems, or class-balanced undersampling which can result in information loss. In this paper, we present the TRAS (TRAnsfer and Share) to effectively utilize long-tailed semi-supervised data. TRAS transforms the imbalanced pseudo-label distribution of a traditional SSL model via a delicate function to enhance the supervisory signals for minority classes. It then transfers the distribution to a target model such that the minority class will receive significant attention. Interestingly, TRAS shows that more balanced pseudo-label distribution can substantially benefit minority-class training, instead of seeking to generate accurate pseudo-labels as in previous works. To simplify the approach, TRAS merges the training of the traditional SSL model and the target model into a single procedure by sharing the feature extractor, where both classifiers help improve the representation learning. According to extensive experiments, TRAS delivers much higher accuracy than state-of-the-art methods in the entire set of classes as well as minority classes. Code for TRAS is available at https://github.com/Stomach-ache/TRAS. [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 08856125 |
| DOI: | 10.1007/s10994-022-06247-z |
