Towards Human-Centered and Efficient Video Synthesis: A Survey of Multimodal Diffusion Models

Abstract excerpt

<title>Abstract</title> <p>Multimodal video diffusion models have emerged as transformative tools for controlled video synthesis, integrating text, images, audio, and pose sequences to generate semantically meaningful content. Despite significant advances, critical gaps persist in temporal consistency, multimodal alignment, and human-centric motion generation. Existing surveys have not addressed clearly the compl...

Topics

Open a Topic to create a Post that cites this publication.

• Algorithms
• Deep Learning
• Image Enhancement
• Image Processing, Computer-Assisted
• Phenotype
• Retina

Identifiers and source

Literature Corpus work

932cda94-5921-5433-a4cd-ed9a491129e6

DOI

10.21203/rs.3.rs-7533477/v1