CodeCap-Bench evaluates the effectiveness of code subtitles
The story so far
CodeCap-Bench demonstrates that code subtitles have superior reasoning accuracy and reconstruction fidelity compared to natural language subtitles in video question answering and reconstruction tasks. Mainstream closed-source MLLMs perform well in code subtitle generation but struggle to capture complex temporal dynamics, with significant gaps in open-source models. CodeCap-Bench includes 323 structured 2D content videos covering five categories and uses fine-grained annotation and a multi-level evaluation framework for assessment.
AI-written from the reports · Updated 22 days ago
Coverage timeline
Follow the coverage to see every side of the story.
- Hugging Face Daily PapersCodeCap-Bench: Can Code Serve as a Good Video Caption?
CodeCap-Bench demonstrates that code captions outperform natural language captions in terms of reasoning accuracy and reconstruction fidelity on video question answering and reconstruction tasks. The benchmark includes 323 structured 2D content videos, covering five categories. Evaluation finds that mainstream closed-source MLLMs perform well in code caption generation but struggle to capture complex temporal dynamics, while open-source models have a significant gap.