AI Paper Atlas

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

Fighting Fire with Fire: Contrastive Debiasing without Bias-free Data via Generative Bias-transformation

Yeonsung Jung, Hajin Shim, June Yong Yang, Eunho Yang

ICML 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Data augmentation and generalization bias · 0 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 bias-transformed views with contrastive learning instead of a vanilla classifierCDvG · Colored MNIST, 0.0% bias-freeaccuracy: 12.53 -> 96.16Improved
Combine CDvG with LfF sample weightingCDvG · Corrupted CIFAR-10, 0.5% bias-freeaccuracy: 31.50 -> 39.07Improved
Use CDvG rather than the strongest listed baselineBiased MNIST · Biased MNISTaccuracy: 48.6 -> 49.48Improved
Use CDvG+LfF rather than LfFImageNet-9 · IN-aaccuracy: 29.4 -> 34.6Improved
Use CDvG+LfF rather than LfFWaterbirds, ResNet-18 · Waterbirds, worst-groupaccuracy: 68.02 -> 74.92Improved
Add the bias-transformation generator G (ablation)CDvG on Corrupted CIFAR-10 · Corrupted CIFAR-10, 0.0% bias-freeaccuracy: 18.29 -> 28.44Improved
Add the bias-transformation generator G (ablation)CDvG on Corrupted CIFAR-10 · Corrupted CIFAR-10, 0.5% bias-freeaccuracy: 20.75 -> 31.50Improved
Add the bias-transformation generator G (ablation)CDvG on Corrupted CIFAR-10 · Corrupted CIFAR-10, 1.0% bias-freeaccuracy: 21.37 -> 33.25Improved
Add the bias-transformation generator G (ablation)CDvG on Corrupted CIFAR-10 · Corrupted CIFAR-10, 2.0% bias-freeaccuracy: 24.50 -> 35.16Improved
Add the bias-transformation generator G (ablation)CDvG on Corrupted CIFAR-10 · Corrupted CIFAR-10, 5.0% bias-freeaccuracy: 31.13 -> 42.75Improved
Add contrastive loss (ablation)CDvG on Corrupted CIFAR-10 · Corrupted CIFAR-10, 0.0% bias-freeaccuracy: 24.25 -> 28.44Improved
Add contrastive loss (ablation)CDvG on Corrupted CIFAR-10 · Corrupted CIFAR-10, 0.5% bias-freeaccuracy: 26.24 -> 31.50Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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