The paper introduces a framework to tackle modality imbalance in multimodal learning by focusing on sample-level variations. It defines a Modality Gap metric to measure prediction discrepancies, models the resulting bimodal distribution with a Gaussian Mixture Model, and uses Bayesian probabilities for soft separation of balanced and imbalanced samples. A two‑stage training process—Warm‑up and Adaptive Training—reallocates loss weights based on the GMM, strengthening alignment for imbalanced samples while favoring fusion for balanced ones, and shows superior performance over existing baselines.
arXiv:2607. 05019v1 Announce Type: new Abstract: In multimodal classification, late-fusion approaches classify concatenated modality-specific features extracted by unimodal neural networks.
arXiv:2608. 02769v1 Announce Type: cross Abstract: Multimodal supervised learning seeks to leverage multiple heterogeneous data sources to improve predictive performance.
arXiv:2607. 27289v1 Announce Type: new Abstract: The promise of multimodal fusion lies in combining complementary sources of evidence, yet more evidence does not always yield a better prediction.
arXiv:2606. 11190v1 Announce Type: new Abstract: Cross-modal alignment (CA) and cross-modal prediction (CP) are the dominant paradigms for multimodal representation learning, yet there is no systematic understanding of when each succeeds, when each fails, and when cross-modal training helps at all -- a gap that leaves practitioners, especially in scientific domains like biomedicine or astrophysics, with heterogeneous instruments and multiple levels of organization and measurement, unable to diagnose why standard methods underperform the best single modality.
InfMasking introduces a unified multimodal learning framework that uses Infinite Masking to encourage fusion modules to capture synergistic information across modalities. By randomly masking modality features before fusion and enforcing consistency with an unmasked reference, the method discourages reliance on dominant or shared signals and promotes meaningful cross‑modal interactions. The approach models masked fused representations as Gaussian, optimizing closed‑form variational bounds, and demonstrates improved predictive performance and higher synergy metrics on benchmark datasets.
Cross-modal alignment (CA) and cross-modal prediction (CP) are the dominant paradigms for multimodal representation learning, yet there is no systematic understanding of when each succeeds, when each fails, and when cross-modal training helps at all -- a gap that leaves practitioners, especially in scientific domains like biomedicine or astrophysics, with heterogeneous instruments and multiple levels of organization and measurement, unable to diagnose why standard methods underperform the best single modality. We develop a unified linear framework that addresses both questions.
The paper introduces a unified exponential framework for generalising K‑fold cross‑validation (CV) accuracy into conservative risk bounds, called gamma‑CUBV. It models dependence between folds via a joint sub‑Gaussian proxy matrix, yielding an effective number of folds and showing that more folds do not always increase evidence when data are strongly correlated. The framework extends to posterior predictor distributions and weighted multi‑source fusion, and is empirically evaluated on linear classifiers trained on heterogeneous multimodal Gaussian mixtures, demonstrating that corrected resubstitution can be tighter than K‑fold CV in low‑sample, high‑heterogeneity settings.
arXiv:2606. 15743v1 Announce Type: new Abstract: This paper addresses the missing-modality challenge in multi-modal learning by introducing Unsupervised Learning for Missing Modalities in Multi-Modal Learning (UL4M4), a flexible framework that imputes missing feature embeddings in a task-independent manner before supervised prediction.
The paper introduces Fusion Anything Model (FAM), a foundation model designed for generalized multimodal data fusion that can handle arbitrary modality combinations and prediction tasks. FAM is trained on large-scale synthetic multimodal datasets generated via Structural Multimodal Causal Models (SMCMs), enabling it to encode transferable multimodal correlations. Experiments on 18 real-world datasets across 12 modalities and 11 tasks show that FAM performs competitively with specialized models without requiring task-specific adaptation.
arXiv:2609.23533v1 Announce Type: new Abstract: Multimodal classifiers can converge to modality-dominant solutions in which one modality dominates the joint prediction, suppressing the learning of ot...
The paper introduces a generalized multimodal foundation model that can handle arbitrary combinations of modalities and prediction tasks. It trains on large-scale synthetic multimodal datasets with diverse causal structures to learn transferable multimodal correlations. Experiments on 18 real-world datasets across 12 modalities and 11 tasks show competitive performance compared to specialized models without task-specific adaptation.
The paper identifies that in multimodal learning, optimization often produces asymmetric certainty gains, with the stronger modality becoming more confident than the weaker one, which leads to imbalanced contributions and suboptimal performance. The authors attribute this issue to unimodal characteristics and propose a Max Confidence Regularization (MaxCR) method that tracks each modality’s semantic confidence via a nonlinear sparsity measure and applies max suppression and excitation to balance confidence levels. Experiments on standard datasets demonstrate that MaxCR improves overall performance compared to state‑of‑the‑art multimodal baselines.