search

svip-lab / impersonator

PyTorch implementation of our ICCV 2019 paper: Liquid Warping GAN: A Unified Framework for Human Motion Imitation, Appearance Transfer and Novel View Synthesis
https://svip-lab.github.io/project/impersonator
Other
1.73k stars 318 forks source link

still generating low resolution images after training on high resolution data #45

Open OldChi opened 4 years ago

OldChi commented 4 years ago

hi I trained on your 10241024 data and set image size to 512 during training. however, using the result, I still generated 256256 images, how can i modify the code so that i can generate 512*512 images?

StevenLiuWen commented 4 years ago

Hi, @OldChi , You need to set the flag --image_size 512 to generate 512 x 512 resolution images when you have trained the model on such resolution.

MockerPeking commented 4 years ago

I‘ve checked all the issues, wondering what is the highest resolution image that the Motion Imitation part of this project can generate？1024×1024？ Which type of graphics card is generally required？@StevenLiuWen

piaozhx commented 4 years ago

@MockerPeking if you want to train a high resolution video (like 1024x1024), large gpu memory is necessary, like rtx titan (24g) p40 (24g) v100(32g),