AI Paper Atlas

Research map / Label-efficient and robust learning / Domain adaptation and generalization

MER-DG: Modality-Entropy Regularization for Multimodal Domain Generalization

Yavuz Yarici, Ghassan AlRegib

ICML 2026 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Optimization-based domain generalization · 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
add MER-DG entropy regularizer to standard fusionFusion (video+audio) · EPIC-Kitchens multi-sourceavg top-1 accuracy: 59.09 -> 62.82Improved
add MER-DG to FusionFusion (video+audio) · HAC multi-sourceavg top-1 accuracy: 63.70 -> 71.43Improved
add MER-DG to CMRFCMRF (video+audio) · EPIC-Kitchens multi-sourceavg top-1 accuracy: 63.91 -> 64.60Improved
add MER-DG to CMRFCMRF (video+flow) · HAC multi-sourceavg top-1 accuracy: 72.64 -> 72.72No clear effect
add marginal-entropy loss only (ablation)MER-DG on Fusion · EPIC-Kitchens V+Aavg accuracy: 59.09 -> 62.55Improved
add spectral-entropy loss only (ablation)MER-DG on Fusion · EPIC-Kitchens V+Aavg accuracy: 59.09 -> 61.19Improved
combine marginal and spectral entropy losses (ablation)MER-DG · EPIC-Kitchens V+Aavg accuracy: 62.55 -> 62.82Improved
add MER-DG to independently trained unimodal encoder (ablation)Uni-Video · EPIC-Kitchens target domainaccuracy: 59.46 -> 59.51No clear effect
add MER-DG to fusion trainingFusion video encoder · EPIC-Kitchens target domainstandalone accuracy: 56.53 -> 58.61Improved
use cross-attention fusion with MER-DG instead of late fusion (ablation)MER-DG · EPIC-Kitchens V+Aavg accuracy: 59.48 -> 61.13Improved
use standard regularizers (label smoothing, weight decay, dropout, noise) instead of MER-DG (ablation)Fusion · EPIC-Kitchens V+Aavg accuracy: 59.09 -> 60.21 (best, label smoothing)No clear effect
add MER-DG for robustness to Gaussian noise or missing audioFusion · EPIC-Kitchens V+A corruptionsaccuracy drop: -10.94 -> -8.55 (drop audio)Improved

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