nihaomiao / CVPR23_LFDM

The pytorch implementation of our CVPR 2023 paper "Conditional Image-to-Video Generation with Latent Flow Diffusion Models"
BSD 2-Clause "Simplified" License
443 stars 42 forks source link

result in new people is bad #31

Closed chenbolin-master closed 1 year ago

chenbolin-master commented 1 year ago

result in new people is bad. 11111

nihaomiao commented 1 year ago

For the unseen person from the new domain, it is reasonable that the performance will degrade.