期刊
2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022)
卷 -, 期 -, 页码 20792-20802出版社
IEEE COMPUTER SOC
DOI: 10.1109/CVPR52688.2022.02016
关键词
-
资金
- ASU
- Mayo Clinic
- NIH [R01HL128785]
- National Science Foundation (NSF) [ACI-1548562]
This study presents a framework called DiRA that combines discriminative, restorative, and adversarial learning for self-supervised learning in computer vision and medical imaging. The experiments show that DiRA produces more generalizable representation, reduces annotation cost, and improves lesion localization accuracy.
Discriminative learning, restorative learning, and adversarial learning have proven beneficial for self-supervised learning schemes in computer vision and medical imaging. Existing efforts, however, omit their synergistic effects on each other in a ternary setup, which, we envision, can significantly benefit deep semantic representation learning. To realize this vision, we have developed DiRA, the first framework that unites discriminative, restorative, and adversarial learning in a unified manner to collaboratively glean complementary visual information from unlabeled medical images for fine-grained semantic representation learning. Our extensive experiments demonstrate that DiRA (1) encourages collaborative learning among three learning ingredients, resulting in more generalizable representation across organs, diseases, and modalities; (2) outperforms fully supervised ImageNet models and increases robustness in small data regimes, reducing annotation cost across multiple medical imaging applications; (3) learns fine-grained semantic representation, facilitating accurate lesion localization with only image-level annotation; and (4) enhances state-of-the-art restorative approaches, revealing that DiRA is a general mechanism for united representation learning. All code and pretrained models are available at https://github.com/JLiangLab/DiRA.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据