Verified80% confidenceFactTime unknown
Latent Diffusion通过VAE将图像从512×512×3的像素空间压缩到64×64×4的潜空间表示，数据量缩减约48倍
7
Sources
80%
Confidence
Long-term
Relevance
6/2/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedLDM通过预训练的KL-VAE将512×512 RGB图像从786,432维压缩至64×64×4的潜在空间，维度压缩比约48倍79% similarVerifiedLTX compresses video into a low-dimensional latent space via a VAE (typically by a factor of 8 to 16) before running the diffusion process, in order to reduce computational cost.76% similarUnverifiedLDM将图像尺寸通常压缩8倍（如512×512压缩为64×64的潜在表示），计算量降低数十倍，使消费级显卡运行成为可能71% similarUnverifiedTraditional diffusion models process approximately 780,000 values per step in a 512×512×3 pixel space.67% similarUnverifiedJPEG压缩中quality参数本质上控制的是量化步长，值越低丢弃的高频信息越多，文件越小但细节损失越大67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/36456API
curl https://kongchang.com/api/v1/knowledge/claims/36456MCP
get_claim(id=36456)