AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Model Doctor: A Simple Gradient Aggregation Strategy for Diagnosing and Treating CNN Classifiers

Zunlei Feng, Jiacong Hu, Sai Wu, XiaoTian Yu, Jie Song, Mingli Song

AAAI 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 11 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 and space-wise constraintsVGG-16 · mini-ImageNet (100 categories)accuracy: +5.42Improved
Add the space-wise constraintVGG-16 · mini-ImageNetaccuracy: +5.27Improved
Add the channel-wise constraintVGG-16 · mini-ImageNetaccuracy: +4.75Improved
Use 12% annotated samples for the space-wise constraint (ablation)VGG-16 +Space · mini-ImageNetaccuracy increase: +7.53%Improved
Use 0-pixel expansion of the background mask (ablation)VGG-16 +Space · mini-ImageNetaccuracy increase: -7.48%Hurt
Use 30-pixel expansion of the background mask (ablation)VGG-16 +Space · mini-ImageNetaccuracy increase: +5.58%Improved
Constrain deeper rather than shallow layers (ablation)Model Doctor on VGG-16 · mini-ImageNetaccuracy increaseImproved

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