Flow-Style-VTON

Flow-Style-VTON copied to clipboard

tengolagan

Hi, why is the nb of output channels of the StyledConvBlock is set to 49 by default ? thanks

tengolagan

When I try to train PB and PF models, I see the need to train 2 stages for each model, stage 2 (e2e) will take the checkpoint of the warp...

zero-nnkn

Can you help explain the modulation convolution in the paper? I don't know much about that.

lujinglmy

get bad results

2

Hi, Thanks for sharing your work. I got bad results when running test.py with VTON test dataset and the non_augmented checkpoint.   Almost all the pictures got bad results...

LSY1111

Thank you for your excellent work, I would like to know how long the training time is in total, if it is too long, how can I reduce the training...

changshuowang

Python built-in function for extracting the cloth's edge image

1

Hello, What is the built-in function you have used for extracting the cloth's edge image ? https://github.com/SenHe/Flow-Style-VTON/blob/ed25d525a53fadaf3ea794d0beb6affbb75f8160/test/test.py#L75 Thanks, Surendra

onchiptech

Hello, First, thank you for sharing your great work! I have a question as below: The torchvision automatically download a VGG-19 ckpt, but why should I downloaded the pretrained VGG-19...

SHShim0513

训练检测点

4

Hello, where can I get the training model for each stage

cskyy

Please make sure you have read and understood all the code and then ask this question. _Originally posted by @SenHe in https://github.com/SenHe/Flow-Style-VTON/issues/20#issuecomment-1207210638_

cskyy