Evaluating and Improving Domain Invariance in Contrastive Self-Supervised Learning by Extrapolating the Loss Function

Evaluating and Improving Domain Invariance in Contrastive Self-Supervised Learning by Extrapolating the Loss Function

Despite the remarkable progress of self-supervised learning (SSL), how self-supervised representations generalize to out-of-distribution data remains little understood. In this paper, we study the effects of distribution shifts on self-supervised representations. Our findings indicate that self-supe...

Full description

Bibliographic Details
Main Authors: Samira Zare, Hien Van Nguyen
Format: Article
Language:English
Published: IEEE 2023-01-01
Series:IEEE Access
Subjects:
Self-supervised learning
invariant learning
robust optimization
out-of-distribution generalization
Online Access:https://ieeexplore.ieee.org/document/10345630/

Internet

https://ieeexplore.ieee.org/document/10345630/

Similar Items