AI Paper Atlas

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

Improving Contrastive Learning on Imbalanced Data via Open-World Sampling

Ziyu Jiang, Tianlong Chen, Ting Chen, Zhangyang Wang

NeurIPS 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Contrastive learning for imbalanced data · 12 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
Use MAK to select 10K external imagesSimCLR, ResNet-50 on ImageNet-100-LT · ImageNet-900; linear separabilityAll accuracy: 71.2±0.2 -> 72.7±0.4Improved
Use MAK rather than random samplingSimCLR, ResNet-50 on ImageNet-100-LT · ImageNet-900; linear separabilityStd (imbalancedness): 3.5±0.5 -> 3.0±0.1Improved
Use MAK to select 10K external imagesSimCLR, ResNet-50 on ImageNet-100-LT · ImageNet-900; few-shotAll accuracy: 51.1±0.1 -> 51.9±0.4Improved
Use MAK rather than random samplingSimCLR, ResNet-50 on ImageNet-100-LT · ImageNet-Places-Mix; linear separabilityAll accuracy: 69.8±0.5 -> 71.1±0.5Improved
Use MAK rather than random samplingSimCLR, ResNet-50 on ImageNet-100-LT · ImageNet-Places-Mix; few-shotAll accuracy: 48.7±0.4 -> 49.3±0.7Improved
Add tailness selection alone (ablation)MAK sampling on ImageNet-100-LT · ImageNet-900; linear separabilityAll accuracy: 71.2±0.2 -> 70.9±0.4Hurt
Add proximity selection alone (ablation)MAK sampling on ImageNet-100-LT · ImageNet-900; linear separabilityAll accuracy: 71.2±0.2 -> 70.2±0.4Hurt
Add diversity selection alone (ablation)MAK sampling on ImageNet-100-LT · ImageNet-900; linear separabilityAll accuracy: 71.2±0.2 -> 70.0±0.4Hurt
Combine tailness and proximity (ablation)MAK sampling on ImageNet-100-LT · ImageNet-900; linear separabilityAll accuracy: 71.2±0.2 -> 72.2±0.2Improved
Add diversity to tailness and proximity (ablation)MAK sampling on ImageNet-100-LT · ImageNet-900; linear separabilityAll accuracy: 72.2±0.2 -> 72.7±0.4Improved
Remove tailness from full MAK (ablation)MAK sampling on ImageNet-100-LT · ImageNet-900; few-shotAll accuracy: 51.9±0.4 -> 49.1±0.3Hurt
Remove proximity from full MAK (ablation)MAK sampling on ImageNet-100-LT · ImageNet-900; few-shotAll accuracy: 51.9±0.4 -> 49.3±0.4Hurt

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