Proposes a plug-and-play multimodal diffusion guidance mechanism that improves video-to-audio generation quality and alignment by enforcing unified multimodal coherence on pretrained audio diffusion models.
Gramian multimodal representation learning and alignment,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Training-Free Multimodal Guidance for Video to Audio Generation
Proposes a plug-and-play multimodal diffusion guidance mechanism that improves video-to-audio generation quality and alignment by enforcing unified multimodal coherence on pretrained audio diffusion models.