[2506.00835] SynPO: Synergizing Descriptiveness and Preference Optimization for Video Detailed Captioning

Computer Science > Artificial Intelligence arXiv:2506.00835 (cs) [Submitted on 1 Jun 2025 (v1), last revised 22 Mar 2026 (this version, v2)] Title:SynPO: Synergizing Descriptiveness and Preference Optimization for Video Detailed Captioning Authors:Jisheng Dang, Yizhou Zhang, Hao Ye, Teng Wang, Siming Chen, Huicheng Zheng, Yulan Guo, Jianhuang Lai, Bin Hu View a PDF of the paper titled SynPO: Synergizing Descriptiveness and Preference Optimization for Video Detailed Captioning, by Jisheng Dang and 8 other authors View PDF HTML (experimental) Abstract:Fine-grained video captioning aims to generate detailed, temporally coherent descriptions of video content. However, existing methods struggle to capture subtle video dynamics and rich detailed information. In this paper, we leverage preference learning to enhance the performance of vision-language models in fine-grained video captioning, while mitigating several limitations inherent to direct preference optimization (DPO). First, we propose a pipeline for constructing preference pairs that leverages the intrinsic properties of VLMs along with partial assistance from large language models, achieving an optimal balance between cost and data quality. Second, we propose Synergistic Preference Optimization (SynPO), a novel optimization method offering significant advantages over DPO and its variants. SynPO prevents negative preferences from dominating the optimization, explicitly preserves the model's language capability to avoid d...