Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | https://arxiv.org/abs/2505.18657 |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866908378695991296 |
|---|---|
| author | Zheng, Xu Liao, Chenfei Fu, Yuqian Lei, Kaiyu Lyu, Yuanhuiyi Jiang, Lutao Ren, Bin Chen, Jialei Wang, Jiawen Li, Chengxin Zhang, Linfeng Paudel, Danda Pani Huang, Xuanjing Jiang, Yu-Gang Sebe, Nicu Tao, Dacheng Van Gool, Luc Hu, Xuming |
| author_facet | Zheng, Xu Liao, Chenfei Fu, Yuqian Lei, Kaiyu Lyu, Yuanhuiyi Jiang, Lutao Ren, Bin Chen, Jialei Wang, Jiawen Li, Chengxin Zhang, Linfeng Paudel, Danda Pani Huang, Xuanjing Jiang, Yu-Gang Sebe, Nicu Tao, Dacheng Van Gool, Luc Hu, Xuming |
| contents | Recent advances in Multimodal Large Language Models (MLLMs) have shown promising results in integrating diverse modalities such as texts and images. MLLMs are heavily influenced by modality bias, often relying on language while under-utilizing other modalities like visual inputs. This position paper argues that MLLMs are deeply affected by modality bias. Firstly, we diagnose the current state of modality bias, highlighting its manifestations across various tasks. Secondly, we propose a systematic research road-map related to modality bias in MLLMs. Thirdly, we identify key factors of modality bias in MLLMs and offer actionable suggestions for future research to mitigate it. To substantiate these findings, we conduct experiments that demonstrate the influence of each factor: 1. Data Characteristics: Language data is compact and abstract, while visual data is redundant and complex, creating an inherent imbalance in learning dynamics. 2. Imbalanced Backbone Capabilities: The dominance of pretrained language models in MLLMs leads to overreliance on language and neglect of visual information. 3. Training Objectives: Current objectives often fail to promote balanced cross-modal alignment, resulting in shortcut learning biased toward language. These findings highlight the need for balanced training strategies and model architectures to better integrate multiple modalities in MLLMs. We call for interdisciplinary efforts to tackle these challenges and drive innovation in MLLM research. Our work provides a fresh perspective on modality bias in MLLMs and offers insights for developing more robust and generalizable multimodal systems-advancing progress toward Artificial General Intelligence. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_18657 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MLLMs are Deeply Affected by Modality Bias Zheng, Xu Liao, Chenfei Fu, Yuqian Lei, Kaiyu Lyu, Yuanhuiyi Jiang, Lutao Ren, Bin Chen, Jialei Wang, Jiawen Li, Chengxin Zhang, Linfeng Paudel, Danda Pani Huang, Xuanjing Jiang, Yu-Gang Sebe, Nicu Tao, Dacheng Van Gool, Luc Hu, Xuming Artificial Intelligence Recent advances in Multimodal Large Language Models (MLLMs) have shown promising results in integrating diverse modalities such as texts and images. MLLMs are heavily influenced by modality bias, often relying on language while under-utilizing other modalities like visual inputs. This position paper argues that MLLMs are deeply affected by modality bias. Firstly, we diagnose the current state of modality bias, highlighting its manifestations across various tasks. Secondly, we propose a systematic research road-map related to modality bias in MLLMs. Thirdly, we identify key factors of modality bias in MLLMs and offer actionable suggestions for future research to mitigate it. To substantiate these findings, we conduct experiments that demonstrate the influence of each factor: 1. Data Characteristics: Language data is compact and abstract, while visual data is redundant and complex, creating an inherent imbalance in learning dynamics. 2. Imbalanced Backbone Capabilities: The dominance of pretrained language models in MLLMs leads to overreliance on language and neglect of visual information. 3. Training Objectives: Current objectives often fail to promote balanced cross-modal alignment, resulting in shortcut learning biased toward language. These findings highlight the need for balanced training strategies and model architectures to better integrate multiple modalities in MLLMs. We call for interdisciplinary efforts to tackle these challenges and drive innovation in MLLM research. Our work provides a fresh perspective on modality bias in MLLMs and offers insights for developing more robust and generalizable multimodal systems-advancing progress toward Artificial General Intelligence. |
| title | MLLMs are Deeply Affected by Modality Bias |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2505.18657 |