Unverified85% confidenceFactTime unknown
2023年后，以Stable Video Diffusion、Sora为代表的时序扩散模型开始将视频理解为时空连续体，能够在潜空间中同时建模多帧之间的运动关系
1
Sources
85%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedSora adopts a Video Diffusion Transformer (DiT), compressing video into spatiotemporal patches before modeling the diffusion process.77% similarUnverifiedCurrent mainstream text-to-video models are generally based on diffusion model architectures, generating coherent video frame sequences by extending image generation capability along the temporal dimension.76% similarUnverifiedAI video generation tools like Jimeng rely on a combination of Diffusion Models and video temporal consistency techniques.76% similarUnverifiedState-of-the-art video generation models such as Sora and Stable Video Diffusion both adopt DiT or its variants75% similarVerifiedThe video diffusion model adopts a Diffusion Transformer (DiT) architecture and models both spatial and temporal dimensions simultaneously by introducing a 3D spatiotemporal attention mechanism in the temporal attention layers.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/18850API
curl https://kongchang.com/api/v1/knowledge/claims/18850MCP
get_claim(id=18850)