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

arXiv: https://arxiv.org/abs/2605.01967 · HTML: https://mlsanity.com/paper/2605.01967/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 28 references that are in the map.
- First posted on arXiv 3 May 2026.

## Builds on (most cited first, 24 of 24)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [The Kinetics Human Action Video Dataset](https://mlsanity.com/paper/1705.06950/) (2017)
- [SlowFast Networks for Video Recognition](https://mlsanity.com/paper/1812.03982/) (2018)
- [Barlow Twins: Self-Supervised Learning via Redundancy Reduction](https://mlsanity.com/paper/2103.03230/) (2021)
- [Similarity of Neural Network Representations Revisited](https://mlsanity.com/paper/1905.00414/) (2019)
- [VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learning](https://mlsanity.com/paper/2105.04906/) (2021)
- [Multimodal Machine Learning: A Survey and Taxonomy](https://mlsanity.com/paper/1705.09406/) (2017)
- [Scaling Egocentric Vision: The EPIC-KITCHENS Dataset](https://mlsanity.com/paper/1804.02748/) (2018)
- [VGGSound: A Large-scale Audio-Visual Dataset](https://mlsanity.com/paper/2004.14368/) (2020)
- [ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation](https://mlsanity.com/paper/1811.12833/) (2018)
- [Whitening for Self-Supervised Representation Learning](https://mlsanity.com/paper/2007.06346/) (2020)
- [Domain Generalization for Medical Imaging Classification with Linear-Dependency Regularization](https://mlsanity.com/paper/2009.12829/) (2020)
- [Invariant Information Bottleneck for Domain Generalization](https://mlsanity.com/paper/2106.06333/) (2021)
- [SimMMDG: A Simple and Effective Framework for Multi-modal Domain Generalization](https://mlsanity.com/paper/2310.19795/) (2023)
- [Domain Generalization through Audio-Visual Relative Norm Alignment in First Person Action Recognition](https://mlsanity.com/paper/2110.10101/) (2021)
- [Self-Supervised Learning via Maximum Entropy Coding](https://mlsanity.com/paper/2210.11464/) (2022)
- [Unexplored Faces of Robustness and Out-of-Distribution: Covariate Shifts in Environment and Sensor Domains](https://mlsanity.com/paper/2404.15882/) (2024)
- [Bridging Domain Generalization to Multimodal Domain Generalization via Unified Representations](https://mlsanity.com/paper/2507.03304/) (2025)
- [An Information Maximization Based Blind Source Separation Approach for Dependent and Independent Sources](https://mlsanity.com/paper/2205.00794/) (2022)
- [Hierarchical and Multimodal Data for Daily Activity Understanding](https://mlsanity.com/paper/2504.17696/) (2025)
- [Exploiting the Distortion-Semantic Interaction in Fisheye Data](https://mlsanity.com/paper/2305.00079/) (2023)
- [Modality-Balanced Collaborative Distillation for Multi-Modal Domain Generalization](https://mlsanity.com/paper/2511.20258/) (2025)
- [AdaDim: Dimensionality Adaptation for SSL Representational Dynamics](https://mlsanity.com/paper/2505.12576/) (2025)
- [Subject Invariant Contrastive Learning for Human Activity Recognition](https://mlsanity.com/paper/2507.03250/) (2025)

## Abstract (arXiv)

Deploying multimodal models in real-world scenarios requires generalization to new environments where recording conditions differ from training, a challenge known as multimodal domain generalization (MMDG). Standard architectures employ separate encoders for each modality and a fusion module, training the system end-to-end by optimizing on the fused features. In this paper, we identify that such joint optimization causes encoders to exploit cross-modal co-occurrences, statistical relationships between modalities that arise from source-specific recording conditions, rather than learning domain-invariant features. We term this failure mode Fusion Overfitting. To address this, we propose Modality-Entropy Regularization for Domain Generalization (MER-DG), which maximizes the entropy of each encoder's feature distribution to preserve feature diversity. MER-DG is architecture-agnostic and integrates into existing multimodal frameworks as an additive loss term. Extensive experiments on EPIC-Kitchens and HAC benchmarks demonstrate average improvements of approximately 5% over standard fusion and approximately 2% over state-of-the-art methods.