Instance-dependent label noise learning via separating style from content

Han-Wen Deng

doi:10.1016/j.patrec.2025.04.039

Instance-dependent label noise learning via separating style from content

2025-05-31

0

PRE

AI

DOI：10.1016/j.patrec.2025.04.039

Original

Original request for help

Save

Abstract

En 中文

• Separates content and style to handle instance-dependent label noise. • Combines generative modeling with semi-supervised learning for noisy labels. • Outperforms state-of-the-art on noisy datasets like CIFAR-0 and Clothing1M.

Instance-dependent label noise learning via separating style from content

Abstract

Journal

Organization