MCE: Towards a General Framework for Handling Missing Modalities under Imbalanced Missing Rates

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Zhao, Binyu, Zhang, Wei, Zou, Zhaonian
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866914143256182784
author Zhao, Binyu
Zhang, Wei
Zou, Zhaonian
author_facet Zhao, Binyu
Zhang, Wei
Zou, Zhaonian
contents Multi-modal learning has made significant advances across diverse pattern recognition applications. However, handling missing modalities, especially under imbalanced missing rates, remains a major challenge. This imbalance triggers a vicious cycle: modalities with higher missing rates receive fewer updates, leading to inconsistent learning progress and representational degradation that further diminishes their contribution. Existing methods typically focus on global dataset-level balancing, often overlooking critical sample-level variations in modality utility and the underlying issue of degraded feature quality. We propose Modality Capability Enhancement (MCE) to tackle these limitations. MCE includes two synergistic components: i) Learning Capability Enhancement (LCE), which introduces multi-level factors to dynamically balance modality-specific learning progress, and ii) Representation Capability Enhancement (RCE), which improves feature semantics and robustness through subset prediction and cross-modal completion tasks. Comprehensive evaluations on four multi-modal benchmarks show that MCE consistently outperforms state-of-the-art methods under various missing configurations. The final published version is now available at https://doi.org/10.1016/j.patcog.2025.112591. Our code is available at https://github.com/byzhaoAI/MCE.
format Preprint
id arxiv_https___arxiv_org_abs_2510_10534
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MCE: Towards a General Framework for Handling Missing Modalities under Imbalanced Missing Rates
Zhao, Binyu
Zhang, Wei
Zou, Zhaonian
Computer Vision and Pattern Recognition
Machine Learning
Multimedia
Multi-modal learning has made significant advances across diverse pattern recognition applications. However, handling missing modalities, especially under imbalanced missing rates, remains a major challenge. This imbalance triggers a vicious cycle: modalities with higher missing rates receive fewer updates, leading to inconsistent learning progress and representational degradation that further diminishes their contribution. Existing methods typically focus on global dataset-level balancing, often overlooking critical sample-level variations in modality utility and the underlying issue of degraded feature quality. We propose Modality Capability Enhancement (MCE) to tackle these limitations. MCE includes two synergistic components: i) Learning Capability Enhancement (LCE), which introduces multi-level factors to dynamically balance modality-specific learning progress, and ii) Representation Capability Enhancement (RCE), which improves feature semantics and robustness through subset prediction and cross-modal completion tasks. Comprehensive evaluations on four multi-modal benchmarks show that MCE consistently outperforms state-of-the-art methods under various missing configurations. The final published version is now available at https://doi.org/10.1016/j.patcog.2025.112591. Our code is available at https://github.com/byzhaoAI/MCE.
title MCE: Towards a General Framework for Handling Missing Modalities under Imbalanced Missing Rates
topic Computer Vision and Pattern Recognition
Machine Learning
Multimedia
url https://arxiv.org/abs/2510.10534