Research map / Video / Video generation
MAVIN: Multi-Shot Audio-Visual Generation with Customized Narrative Control
Kaiqi Liu, Yunyao Mao, Ziqi Cai, Zheng Geng, Jing Wang, Qiulin Wang, Xintao Wang, Pengfei Wan, Kun Gai, Shuchen Weng, Boxin Shi
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 boundary-aware attention, hierarchical captions and ID-aware propagation to OVI backbone | MAVIN · MAVINSet 1K | STA: 0.4992 -> 0.9897 | Improved |
| same, versus OVI on video quality | MAVIN · MAVINSet 1K | FVD: 318.4 -> 231.6 | Improved |
| add boundary-aware token routing (ablation) | MAVIN · MAVINSet 1K (with refs) | STA: 0.5580 -> 0.9886 | Improved |
| add boundary-aware token routing (ablation) | MAVIN · MAVINSet 1K (with refs) | CISC: 0.5520 -> 0.6397 | Improved |
| add identity-aware mask (ablation) | MAVIN · MAVINSet 1K (with refs) | CISC: 0.5710 -> 0.6397 | Improved |
| add pretrained timbre encoder (ablation) | MAVIN · MAVINSet 1K (with refs) | Sync: 6.010 -> 6.057 | Improved |
| add pretrained timbre encoder (ablation) | MAVIN · MAVINSet 1K (with refs) | FVD: 242.5 -> 241.9 | No clear effect |
| add multi-agent scripting pipeline (ablation) | MAVIN · MAVINSet 1K (with refs) | STA: 0.6230 -> 0.9886 | Improved |
| add multi-agent scripting pipeline (ablation) | MAVIN · MAVINSet 1K (with refs) | TAMS: 0.569 -> 0.793 | Improved |
| MAVIN versus cascaded T2V+V2A on subject consistency | MAVIN · MAVINSet 1K | SC: 0.9656 -> 0.9695 | Improved |
| users prefer MAVIN over baselines | MAVIN · User study | preference %: best baseline AVQ 9.4 vs 36.8 | Improved |
Closest papers by meaning
- Seeing and Hearing: Open-domain Visual-Audio Generation with Diffusion Latent Aligners (CVPR 2024 · similarity 0.86)
- Read, Watch and Scream! Sound Generation from Text and Video (AAAI 2025 · similarity 0.85)
- AutoAD III: The Prequel - Back to the Pixels (CVPR 2024 · similarity 0.85)
- Hallo2: Long-Duration and High-Resolution Audio-Driven Portrait Image Animation (ICLR 2025 · similarity 0.85)
- MMAudio: Taming Multimodal Joint Training for High-Quality Video-to-Audio Synthesis (CVPR 2025 · similarity 0.85)
- ThinkSound: Chain-of-Thought Reasoning in Multimodal LLMs for Audio Generation and Editing (NeurIPS 2025 · similarity 0.85)
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.
