UGC-VideoCaptioner: An Omni UGC Video Detail Caption Model and New Benchmarks
This addresses the lack of omni datasets and lightweight models for fine-grained multimodal video understanding in real-world UGC settings, though it appears incremental in combining existing techniques.
The paper tackles the problem of video captioning for user-generated content by introducing a new benchmark (UGC-VideoCap) with 1000 TikTok videos and 4000 QA pairs that emphasize balanced audio-visual integration, and proposes a 3B parameter model (UGC-VideoCaptioner) that achieves competitive performance through a two-stage training strategy.
Real-world user-generated videos, especially on platforms like TikTok, often feature rich and intertwined audio visual content. However, existing video captioning benchmarks and models remain predominantly visual centric, overlooking the crucial role of audio in conveying scene dynamics, speaker intent, and narrative context. This lack of omni datasets and lightweight, capable models hampers progress in fine grained, multimodal video understanding. To address these challenges, we introduce UGC-VideoCap, a new benchmark and model framework specifically designed for detailed omnimodal captioning of short form user-generated videos. Unlike prior datasets, UGC-VideoCap emphasizes balanced integration of audio and visual modalities, featuring 1000 TikTok videos annotated through a structured three stage human-in-the-loop pipeline covering audio only, visual only, and joint audio visual semantics. The benchmark also includes 4000 carefully crafted QA pairs probing both unimodal and cross modal understanding. Alongside the dataset, we propose UGC-VideoCaptioner(3B), a 3B parameter captioning model distilled from Gemini 2.5 Flash. Using a novel two-stage training strategy supervised fine tuning followed by Group Relative Policy Optimization (GRPO), our approach enables efficient adaptation from limited data while maintaining competitive performance. Together, our benchmark and model offer a high-quality foundation and a data-efficient solution for advancing omnimodal video captioning in unconstrained real-world UGC settings.