AI Paper Atlas

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

Channel-Wise Knowledge Distillation for Dense Prediction

Changyong Shu, Yifan Liu, Jianfei Gao, Zheng Yan, Chunhua Shen

ICCV 2021 · Label-efficient and robust learning › Knowledge distillation · approach: Distillation with causal and adaptive guidance · 520 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 channel-wise distillation on logitsPSPNet-R18 student · Cityscapes val (logits-map distillation)mIoU: 69.10 -> 74.87Improved
add channel-wise distillation on feature mapPSPNet-R18 student · Cityscapes val (feature-map distillation)mIoU: 69.10 -> 74.27Improved
use channel-wise distillation instead of best spatial method ATPSPNet-R18 student · Cityscapes val (logits-map distillation)mIoU: 72.32 -> 74.87Improved
normalize each channel with softmax and use KL instead of L2 without normalization (ablation)CD · Cityscapes val (logits-map distillation)mIoU: 70.83 -> 74.87Improved
use asymmetric KL instead of MSE on channel-normalized maps (ablation)CD · Cityscapes val (logits-map distillation)mIoU: 71.60 -> 74.87Improved
use asymmetric KL instead of symmetric Bhattacharyya distance (ablation)CD · Cityscapes val (logits-map distillation)mIoU: 72.21 -> 74.87Improved
normalize over channels rather than spatial locations (PI) (ablation)CD · Cityscapes val (feature-map distillation)mIoU: 70.02 -> 74.27Improved
swap teacher and student order in KL divergence (ablation)CD · Cityscapes val (logits-map distillation)training convergenceHurt
lower temperature T so only salient pixels are matched (ablation)CD · Cityscapes val (logits-map distillation)mIoUHurt
initialize student with ImageNet weights (ablation)PSPNet-R18 + CD logits · Cityscapes valmIoU: 71.03 -> 75.90Improved
use halved-width student instead of full-width R18 (ablation)PSPNet-R18 + CD logits · Cityscapes val (from scratch)mIoU: 71.03 -> 68.57Hurt
add channel-wise distillation to RetinaNetRetinaNet R50 · COCO 2017 detectionAP: 37.4 -> 40.8Improved

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