Research map / Video / Video-language understanding
AVoCaDO: An Audiovisual Video Captioner Driven by Temporal Orchestration
Xinlong Chen, Yue Ding, Weihong Lin, Jingyun Hua, Linli Yao, Yang Shi, Bozhou Li, Qiang Liu, Yuanxing Zhang, Pengfei Wan, Liang Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add AVoCaDO SFT on 107K temporally aligned audiovisual captions (ablation) | Qwen2.5-Omni-7B · video-SALMONN-2 testset | Total error rate: 57.1 -> 41.4 | Improved |
| add dialogue-based reward R_D in GRPO (ablation) | AVoCaDO-SFT · video-SALMONN-2 testset | Dialogue F1: 74.4 -> 76.5 | Improved |
| add checklist-based reward R_C in GRPO (ablation) | AVoCaDO GRPO with R_D · video-SALMONN-2 testset | Total error rate: 41.3 -> 37.3 | Improved |
| add length-regularized reward R_L in GRPO (ablation) | AVoCaDO GRPO with R_D and R_C · video-SALMONN-2 testset | Repetition collapse (%): 3.9 -> 0.4 | Improved |
| add length-regularized reward R_L in GRPO (ablation) | AVoCaDO GRPO with R_D and R_C · video-SALMONN-2 testset | Total error rate: 37.3 -> 37.3 | No clear effect |
| fine-tune SFT model on the same 2K samples used for GRPO instead of using GRPO (ablation) | AVoCaDO-SFT · video-SALMONN-2 testset | Total error rate: 41.4 -> 43.0 | Hurt |
| add dialogue reward R_D in GRPO (ablation) | AVoCaDO-SFT · Daily-Omni by caption | accuracy: 48.1 -> 49.5 | Improved |
| generate captions jointly from audio and video instead of concatenating separate captions | Gemini-2.5-Pro captions · Daily-Omni pilot, QA from caption | accuracy gain: +15.8% average, +27.8% on AV Event Alignment | Improved |
| build SFT captions with a two-stage (separate then fuse) procedure instead of direct joint captioning (ablation) | Gemini-2.5-Pro SFT data generation · SFT data quality | information coverage | Improved |
Later papers that compared with this method
2 later accepted papers report AVoCaDO as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- DiaDem: Advancing Dialogue Descriptions in Audiovisual Video Captioning for Multimodal Large Language Models (ECCV 2026 · better on 4 of 4 shared results)
- Temporal and Cross-modal Alignment for Enhanced Audiovisual Video Captioning (ECCV 2026 · better on 4 of 4 shared results)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- video-SALMONN-2 — video-SALMONN: Speech-Enhanced Audio-Visual Large Language Models (ICML 2024) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Vript: A Video Is Worth Thousands of Words (NeurIPS 2024 · similarity 0.88)
- OmniCapBench: A Deep-Structured Evaluation Framework for Fine-Grained Audio-Visual Captioning (NeurIPS 2026 · similarity 0.88)
- Empowering LLMs with Pseudo-Untrimmed Videos for Audio-Visual Temporal Understanding (AAAI 2025 · similarity 0.87)
- AuroraCap: Efficient, Performant Video Detailed Captioning and a New Benchmark (ICLR 2025 · similarity 0.86)
- Aligned Better, Listen Better for Audio-Visual Large Language Models (ICLR 2025 · similarity 0.86)
- Video ReCap: Recursive Captioning of Hour-Long Videos (CVPR 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
