From Captions to Rewards (CAREVL): Leveraging Large Language Model Experts for Enhanced Reward Modeling in Large Vision-Language Models
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866916647100481536 |
|---|---|
| author | Dai, Muzhi Sun, Jiashuo Zhao, Zhiyuan Liu, Shixuan Li, Rui Gao, Junyu Li, Xuelong |
| author_facet | Dai, Muzhi Sun, Jiashuo Zhao, Zhiyuan Liu, Shixuan Li, Rui Gao, Junyu Li, Xuelong |
| contents | Aligning large vision-language models (LVLMs) with human preferences is challenging due to the scarcity of fine-grained, high-quality, and multimodal preference data without human annotations. Existing methods relying on direct distillation often struggle with low-confidence data, leading to suboptimal performance. To address this, we propose CAREVL, a novel method for preference reward modeling by reliably using both high- and low-confidence data. First, a cluster of auxiliary expert models (textual reward models) innovatively leverages image captions as weak supervision signals to filter high-confidence data. The high-confidence data are then used to fine-tune the LVLM. Second, low-confidence data are used to generate diverse preference samples using the fine-tuned LVLM. These samples are then scored and selected to construct reliable chosen-rejected pairs for further training. CAREVL achieves performance improvements over traditional distillation-based methods on VL-RewardBench and MLLM-as-a-Judge benchmark, demonstrating its effectiveness. The code will be released soon. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_06260 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | From Captions to Rewards (CAREVL): Leveraging Large Language Model Experts for Enhanced Reward Modeling in Large Vision-Language Models Dai, Muzhi Sun, Jiashuo Zhao, Zhiyuan Liu, Shixuan Li, Rui Gao, Junyu Li, Xuelong Computer Vision and Pattern Recognition Artificial Intelligence Aligning large vision-language models (LVLMs) with human preferences is challenging due to the scarcity of fine-grained, high-quality, and multimodal preference data without human annotations. Existing methods relying on direct distillation often struggle with low-confidence data, leading to suboptimal performance. To address this, we propose CAREVL, a novel method for preference reward modeling by reliably using both high- and low-confidence data. First, a cluster of auxiliary expert models (textual reward models) innovatively leverages image captions as weak supervision signals to filter high-confidence data. The high-confidence data are then used to fine-tune the LVLM. Second, low-confidence data are used to generate diverse preference samples using the fine-tuned LVLM. These samples are then scored and selected to construct reliable chosen-rejected pairs for further training. CAREVL achieves performance improvements over traditional distillation-based methods on VL-RewardBench and MLLM-as-a-Judge benchmark, demonstrating its effectiveness. The code will be released soon. |
| title | From Captions to Rewards (CAREVL): Leveraging Large Language Model Experts for Enhanced Reward Modeling in Large Vision-Language Models |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2503.06260 |