Unverified60% confidenceFactExact time
SVD (Stable Video Diffusion) explored temporal interpolation in latent space to reduce computational overhead.
2
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
8/24/2026
First Seen
Valid until: 11/22/2026
Sources
Related Entities
Related Claims
UnverifiedFVD是FID在视频领域的扩展版本，在计算分布差异时考虑了时间维度的特征69% similarUnverifiedStable Diffusion is based on the Latent Diffusion architecture, which moves computation from pixel space to a compressed latent space, enabling consumer-grade GPUs to run high-quality image generation.69% similarUnverified2023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67% similarUnverifiedTopics covered by the video series include vector spaces, inner products, matrix operations, subspaces, orthogonality, QR decomposition, LU decomposition, eigendecomposition, symmetric matrices and singular value decomposition (SVD).66% similarUnverified均匀采样按固定时间间隔抽帧，实现简单且计算开销可预测，但对视频内容一无所知，容易错失信息密度不均视频的关键瞬间64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/794863API
curl https://kongchang.com/api/v1/knowledge/claims/794863MCP
get_claim(id=794863)