AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

Knowledge Transfer via Dense Cross-Layer Mutual-Distillation

Anbang Yao, Dawei Sun

ECCV 2020 · Label-efficient and robust learning › Knowledge distillation · approach: Distillation with causal and adaptive guidance · 63 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add deep-supervision auxiliary classifiers and dense cross-layer KD to DML (DCM)DML · CIFAR-100, two WRN-28-10top-1 error: 17.89 -> 16.61Improved
add auxiliary classifiers with only the last-layer KD (DML + DS) (ablation)DML · CIFAR-100, two DenseNet-40-12top-1 error: 23.18 -> 23.08No clear effect
add same-staged cross-layer KD (DCM-1) to DML+DS (ablation)DML + DS · CIFAR-100, two DenseNet-40-12top-1 error: 23.18 -> 22.86Improved
add different-staged cross-layer KD (DCM-2) instead of same-staged (ablation)DCM · CIFAR-100, two DenseNet-40-12top-1 error: 22.86 -> 22.43Improved
use both same-staged and different-staged KD (full DCM) rather than DCM-2 only (ablation)DCM · CIFAR-100, two DenseNet-40-12top-1 error: 22.43 -> 22.35Improved
add independent deep supervision only (no mutual distillation)ResNet-164 · CIFAR-100top-1 error: 22.56 -> 21.38Improved
add DCM versus DS alone (ablation)WRN-28-10 · CIFAR-100top-1 error: 18.32 -> 16.61Improved
use 2 auxiliary classifiers instead of 1 (ablation)DCM · CIFAR-100, two WRN-28-10top-1 error: 17.40 -> 16.61Improved
use 3 auxiliary classifiers instead of 2 (ablation)DCM · CIFAR-100, two WRN-28-10top-1 error: 16.61 -> 16.59No clear effect
use default strong auxiliary classifiers instead of simple APFC ones (ablation)DCM · CIFAR-100, DenseNet-40-12 & WRN-28-10top-1 error: 25.10 -> 22.35Improved
use default width auxiliary classifiers instead of narrow ones (ablation)DCM · CIFAR-100, DenseNet-40-12 & WRN-28-10top-1 error: 16.88 -> 16.61Improved
train ResNet-110 with KD from pre-trained WRN-28-10 teacherResNet-110 · CIFAR-100top-1 error: 26.55 -> 26.66No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy