Unverified50% confidenceFactExact time
VAE的编码器将图像压缩到低维潜空间，通常将空间分辨率缩小8倍（1024×1024变为128×128）
1
Sources
50%
Confidence
Long-term
Relevance
8/27/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedStable Diffusion compresses a 512×512×3 image into a 64×64×4 latent representation via VAE, achieving a compression ratio of 48×.73% similarUnverifiedThe LTX series has designed a 3D VAE with joint spatiotemporal compression tailored to video characteristics, typically compressing by 8 to 16 times before running the diffusion process to reduce computational cost.71% similarUnverifiedA raw video of 512×512 resolution and 24 frames has a data volume of approximately 18 MB, which is reduced to approximately 1/96 of the original after VAE compression into the 64×64×4 latent variable space.70% similarUnverified使用DeepSeek生成的Python图片压缩代码，以quality=85压缩一张756,740字节的JPEG图片后，文件缩小到601,168字节，压缩率为20.6%70% similarUnverified双编码器通常将语义压缩为768维向量，该数字来源于BERT-base的隐藏层大小，工业部署中常压缩到128或256维70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/809929API
curl https://kongchang.com/api/v1/knowledge/claims/809929MCP
get_claim(id=809929)