AI Paper Atlas

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

Distilling Knowledge by Mimicking Features

Guohua Wang, Yifan Ge, Jianxin Wu

TPAMI 2021 · Label-efficient and robust learning › Knowledge distillation · approach: Rethinking distillation objectives · 54 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 LSH loss to the l2 feature mimicking loss (ablation)feature mimicking · CIFAR-100 (ResNet50 -> MobileNetV2)top-1 accuracy (%): 65.73 -> 68.99Improved
add LSH loss to the l2 loss (ablation)feature mimicking · CIFAR-100 (vgg13 -> vgg8)top-1 accuracy (%): 72.33 -> 73.68Improved
replace KD logit loss with l2 feature mimicking (ablation)student training · CIFAR-100 (WRN-40-2 -> WRN-16-2)top-1 accuracy (%): 74.92 -> 75.53Improved
add linear embedding layer (2FC) for feature space alignment (ablation)l2 feature mimicking · VOC2007 (ResNet34 -> ResNet18)mAP (%): 88.75 -> 89.98Improved
use student pretrained by KD with 1FC (ablation)l2 feature mimicking · VOC2007 (ResNet34 -> ResNet18)mAP (%): 88.75 -> 90.89Improved
add 2FC to student already pretrained by KD (ablation)l2 feature mimicking · VOC2007 (ResNet34 -> ResNet18)mAP (%): 90.89 -> 90.77Hurt
two-stage finetuning (align embedding first) (ablation)feature mimicking · VOC2007 (ResNet34 -> ResNet18)mAP (%): 89.98 -> 90.59Improved
remove linear embedding layer (1FC) when dims match (ablation)Ours · CIFAR-100 (resnet32x4 -> resnet8x4)top-1 accuracy (%): 73.40 -> 76.75Hurt
bias init 0 vs median (ablation)LSH module · CIFAR-100 (similar arch)top-1 accuracy (%): 75.62 -> 76.04No clear effect
increase beta from 1 to 7 (ablation)Ours · CIFAR-100 (WRN-40-2 -> WRN-16-2)top-1 accuracy (%): 75.33 -> 76.34Mixed
combine Ours with SSKDSSKD · CIFAR-100 (vgg13 -> vgg8)top-1 accuracy (%): 74.90 -> 75.19Improved
add LSH loss to l2 loss on ImageNet (ablation)Ours · ImageNet (ResNet34 -> ResNet18)top-1 error (%): 28.61 -> 28.28Improved

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