search

YanzuoLu / CFLD

[CVPR 2024 Highlight] Coarse-to-Fine Latent Diffusion for Pose-Guided Person Image Synthesis
MIT License
165 stars 11 forks source link

Classifier-free Guidance for training #19

Closed CHNxindong closed 5 months ago

CHNxindong commented 5 months ago

Hi authors. Thanks again for your great work! I have noticed that your methodology incorporates the use of classifier-free guidance both during the training and testing phases. The configuration parameter 'MODEL.U_COND_DOWN_BLOCK_GUIDANCE' has been set to False, which seemingly implies that the model has not undergone training in an unconditional pose setting. I would greatly appreciate clarification on your approach in this regard.

image

image

YanzuoLu commented 5 months ago

This should be set to True. But in fact, there is not much difference whether these two are enabled or not in the experiments we found. Hope this can help.