BriefGPT.xyz
May, 2021
自监督学习中特征解相关化
On Feature Decorrelation in Self-Supervised Learning
HTML
PDF
Tianyu Hua, Wenxiao Wang, Zihui Xue, Yue Wang, Sucheng Ren...
TL;DR
本文研究了自监督表示学习的框架，在此基础上，发现了完全坍塌和维度坍塌的问题，并将维度坍塌与特征装饰相关联以提高表示的稳健性。
Abstract
In
self-supervised representation learning
, a common idea behind most of the state-of-the-art approaches is to enforce the
robustness
of the representations to predefined augmentations. A potential issue of this
→