Unverified70% confidenceFactExact time
FID (Fréchet Inception Distance) evaluates generation quality by comparing the statistical distribution differences between real and generated data in the feature space of a deep neural network, with lower values indicating closer resemblance to real data.
2
Sources
70%
Confidence
Long-term
Relevance
8/7/2026
First Seen
Sources
AI视频生成进化史：从Will Smith吃面翻车到电影级画质
Related Claims
UnverifiedFID（Fréchet Inception Distance）由Heusel等人于2017年提出，用预训练Inception网络提取特征并将分布建模为多元高斯，计算Fréchet距离，FID值越低表明分布越接近77% similarUnverified可借助Fréchet Inception Distance（FID）或Maximum Mean Discrepancy（MMD）等分布距离度量来量化增强后训练分布与真实测试分布之间的偏移74% similarUnverifiedThe core loss functions of contrastive learning (such as InfoNCE and NT-Xent) essentially pull the representations of similar samples closer together while pushing the representations of dissimilar samples further apart.70% similarUnverified量化测量域偏移程度的常用方法包括最大均值差异（MMD）和Fréchet Inception Distance（FID）68% similarUnverified神经网络中各层对精度损失的敏感程度差异显著，patch projection等输入投影层敏感度高，中间主干的注意力与FFN模块对低精度更鲁棒66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/699336API
curl https://kongchang.com/api/v1/knowledge/claims/699336MCP
get_claim(id=699336)