.png){kind=icon}
.png){kind=icon}
Return
PASS: Peer-agreement based sample selection for training with instance dependent noisy labels
.png){kind=icon}
OA
AI
DOI：10.1016/j.imavis.2025.105877.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• We explore the idea that models often disagree on noisy samples and agree on clean samples. • We propose PASS, a clean-sample selection based on 3 models, where two models are used to select clean samples for training the other model. • PASS is highly integrable and can be implemented into other methods. • PASS improves the performances of most learning with noisy label SOTA methods by significant margins.
Keywords:
Noisy-labels
Instance-dependent noise
Noisy-label learning
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
4.2
Papers:
4.0K
Citations:
6.7K
View more organizations