Teach sample-specific knowledge: Separated distillation based on samples

Seonghak Kim

doi:10.1016/j.engappai.2025.112696

Teach sample-specific knowledge: Separated distillation based on samples

2025-10-09

0

PRE

AI

DOI：10.1016/j.engappai.2025.112696

Original

Original request for help

Save

Abstract

En 中文

• We highlight the limitations of previous KD methods based on forward KL divergence. • Datasets are split to tackle mode-averaging and teacher errors in uncertain images. • Correct samples use RKLD loss, while incorrect samples encourage student self-learning. • Our method achieves superior performance in both classification and object detection.

Teach sample-specific knowledge: Separated distillation based on samples

Abstract

Journal

Organization

Cited Papers