AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Frequency-Balanced Retinal Representation Learning with Mutual Information Regularization

Seunghoon Lee, Seongjae Kang, Inhyuk Park, Gitaek Kwon, Jihyeon Baek, Doohyun Park

ICLR 2026 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Masked autoencoder pretraining

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 HighFreqMI regularizer (Lhmi) to MAE reconstruction (ablation)MAE · 5 retinal benchmarks linear probingAUROC: 0.685 -> 0.779Improved
add HighFreqMI regularizer (Lhmi) to MAE reconstruction (ablation)MAE · 5 retinal benchmarks linear probingAUPRC: 0.486 -> 0.614Improved
add HighFreqMI to the auxiliary-loss setting (ablation)MAE + Laux · 5 retinal benchmarks linear probingAUROC: 0.923 -> 0.941Improved
add HighFreqMI to the auxiliary-loss setting (ablation)MAE + Laux · 5 retinal benchmarks linear probingAUPRC: 0.799 -> 0.849Improved
add auxiliary feature alignment (Laux) to MAE (ablation)MAE · 5 retinal benchmarks linear probingAUROC: 0.685 -> 0.923Improved
add Lhmi to MAE with HF masking (ablation)MAE w/ HF masking · 5 retinal benchmarks linear probingAUROC: 0.679 -> 0.737Improved
add Lhmi to MAE with HF input channel (ablation)MAE w/ HF input · 5 retinal benchmarks linear probingAUROC: 0.746 -> 0.769Improved
feed only high-frequency tokens (25%) instead of 75% masked tokens at probingMAE · 5 retinal benchmarks linear probingAUROC: 0.647 -> 0.727Improved
use low-frequency tokens only instead of full inputMAE · 5 retinal benchmarks linear probingAUROC: 0.685 -> 0.641Hurt
set lambda_hmi to 1.0 instead of 0.1 (ablation)RetMAE (no aux) · 5 retinal benchmarks linear probingAUROC: 0.779 -> 0.738Hurt
set lambda_hmi to 0.01 instead of 0.1 (ablation)RetMAE (no aux) · 5 retinal benchmarks linear probingAUROC: 0.779 -> 0.732Hurt
vary lambda_aux among 1.0, 0.1, 0.01 (ablation)RetMAE (aux only) · 5 retinal benchmarks linear probingAUROC: 0.927 / 0.926 / 0.932No clear effect

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