Early Learning Regularization

Introduced by Liu et al. in Early-Learning Regularization Prevents Memorization of Noisy Labels

Source: Early-Learning Regularization Prevents Memorization of Noisy Labels

Paper	Code	Results	Date	Stars

Task	Papers	Share
Active Learning	2	10.53%
Classification	2	10.53%
Memorization	2	10.53%
General Classification	2	10.53%
Pretrained Language Models	1	5.26%
Text Classification	1	5.26%
Medical Image Segmentation	1	5.26%
Semantic Segmentation	1	5.26%
Weakly supervised segmentation	1	5.26%

This feature is experimental; we are continuously improving our matching algorithm.

Component	Type	Add Remove
🤖 No Components Found	You can add them if they exist; e.g. Mask R-CNN uses RoIAlign