# Mitigating Sample-Level Imbalance via Probabilistic Separation for Adaptive Multimodal Fusion

arXiv: https://arxiv.org/abs/2510.21797 · HTML: https://mlsanity.com/paper/2510.21797/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 26 of its 44 references that are in the map.
- First posted on arXiv 20 Oct 2025.

## Builds on (most cited first, 26 of 26)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [VQA: Visual Question Answering](https://mlsanity.com/paper/1505.00468/) (2015)
- [The Kinetics Human Action Video Dataset](https://mlsanity.com/paper/1705.06950/) (2017)
- [Two-Stream Convolutional Networks for Action Recognition in Videos](https://mlsanity.com/paper/1406.2199/) (2014)
- [Multimodal Machine Learning: A Survey and Taxonomy](https://mlsanity.com/paper/1705.09406/) (2017)
- [Look, Listen and Learn](https://mlsanity.com/paper/1705.08168/) (2017)
- [What Makes Training Multi-Modal Classification Networks Hard?](https://mlsanity.com/paper/1905.12681/) (2019)
- [Balanced Multimodal Learning via On-the-fly Gradient Modulation](https://mlsanity.com/paper/2203.15332/) (2022)
- [EPIC-Fusion: Audio-Visual Temporal Binding for Egocentric Action Recognition](https://mlsanity.com/paper/1908.08498/) (2019)
- [Listen to Look: Action Recognition by Previewing Audio](https://mlsanity.com/paper/1912.04487/) (2019)
- [PMR: Prototypical Modal Rebalance for Multimodal Learning](https://mlsanity.com/paper/2211.07089/) (2022)
- [Characterizing and overcoming the greedy nature of learning in multi-modal deep neural networks](https://mlsanity.com/paper/2202.05306/) (2022)
- [Multimodal Representation Learning by Alternating Unimodal Adaptation](https://mlsanity.com/paper/2311.10707/) (2023)
- [Boosting Multi-modal Model Performance with Adaptive Gradient Modulation](https://mlsanity.com/paper/2308.07686/) (2023)
- [MMPareto: Boosting Multimodal Learning with Innocent Unimodal Assistance](https://mlsanity.com/paper/2405.17730/) (2024)
- [ReconBoost: Boosting Can Achieve Modality Reconcilement](https://mlsanity.com/paper/2405.09321/) (2024)
- [On Uni-Modal Feature Learning in Supervised Multi-Modal Learning](https://mlsanity.com/paper/2305.01233/) (2023)
- [Diagnosing and Re-learning for Balanced Multimodal Learning](https://mlsanity.com/paper/2407.09705/) (2024)
- [On-the-fly Modulation for Balanced Multimodal Learning](https://mlsanity.com/paper/2410.11582/) (2024)
- [MMCosine: Multi-Modal Cosine Loss Towards Balanced Audio-Visual Fine-Grained Learning](https://mlsanity.com/paper/2303.05338/) (2023)
- [Speech2Action: Cross-modal Supervision for Action Recognition](https://mlsanity.com/paper/2003.13594/) (2020)
- [AKVSR: Audio Knowledge Empowered Visual Speech Recognition by Compressing Audio Knowledge of a Pretrained Model](https://mlsanity.com/paper/2308.07593/) (2023)
- [Modeling Multimodal Clues in a Hybrid Deep Learning Framework for Video Classification](https://mlsanity.com/paper/1706.04508/) (2017)
- [Simultaneous Detection and Interaction Reasoning for Object-Centric Action Recognition](https://mlsanity.com/paper/2404.11903/) (2024)
- [Cross-Modal Adaptive Dual Association for Text-to-Image Person Retrieval](https://mlsanity.com/paper/2312.01745/) (2023)

## Abstract (arXiv)

Multimodal learning faces modality imbalance, where dominant modalities suppress weaker ones due to inconsistent convergence rates. Existing static or heuristic methods overlook sample-level variations in prediction bias and fail to isolate low-quality outlier samples. To address this, we propose a novel framework to quantitatively diagnose and dynamically mitigate modality imbalance at the sample level. We first introduce a Modality Gap metric to quantify prediction discrepancies between unimodal branches. Empirical analysis reveals a distinct bimodal distribution, reflecting the natural coexistence of balanced and imbalanced sample subgroups. We then employ a Gaussian Mixture Model (GMM) to model this gap distribution, leveraging Bayesian posterior probabilities for probabilistic soft separation of subgroups. Next, we construct a two-stage training framework comprising a Warm-up stage and an Adaptive Training stage. In the Adaptive Training stage, a GMM-guided Adaptive Loss dynamically reallocates optimization priorities, imposing stronger modality alignment penalties on imbalanced samples while prioritizing multimodal fusion for balanced ones. Experimental results demonstrate that our method significantly outperforms current state-of-the-art baselines. Furthermore, fine-tuning on a high-quality balanced subset filtered by the GMM serves as an effective data purification strategy.