Flow-Style-VTON

Flow-Style-VTON copied to clipboard

Error while training on custom dataset

7

Hi, Great work. I am trying to train the model on my custom dataset but I am stuck on this error. File "train_PBAFN_stage1_fs.py", line 106, in concat = torch.cat([preserve_mask.cuda(),densepose,pose.cuda()],1) RuntimeError:...

peach6

Regarding the extraction method for the in-shop Cloth mask, the previous Colab link has become invalid

1

Hello, you previously provided in-cloth mask extracting code based on Unet ([Colab link](https://colab.research.google.com/drive/1HvSB2Mx8WZ-Wo_8FDrc05DaOi4QUE9hG)), but the link is now invalid. Could you please provide another set of related code or reference...

little-misfit

Which dataset should we use to calculate FID score?

2

I read on some resources that the FID score should be calculated using the dataset of generated outputs and the dataset that was used to train the generator. Which datasets...

TianJiaJiaFirst

VITON pre-processing in one colab notebook (just upload your data)

2

I have created a single Colab notebook for VITON pre-processing that allows you to upload your data and quickly process it. This notebook simplifies the pre-processing steps and can save...

MosbehBarhoumi

Hello, I would like to ask how you calculate the FID score? Are you using the try-on results generated from the test and the images of the person inputted? My...

xxxxl888

Hello, I have a problem that the calculated output size is too small (512*1*0）when I train the model, is there anyone who have successfully run the pretrained train_PBAFN_stage1_fs.sh model could give me some suggestions?

13

 )

wyf-gh

Replacement of full body clothing, such as dresses

3

Thank you for doing this great job. I would like to try to get the model to wear a dress. How should I modify the code? Please give me some...

Hanjunzhe

How to test on custom images?

5

Hi, Thankyou for the project. Can you please guide about how to test on our own images?Thankyou

MamoonHaider

support for higher resolution images

2

hi，thanks for great works, i wonder if it can support higher resolution rgb-images input and generate high-resolution results?

TonyFaith

How come ResUnetGenerator doesn't have a method update_learning_rate? Last 3 rows of code in train_PBAFN_e2e_fs.py : Why are we updating model (the warping model) twice ? Shouldn't **model.module.update_learning_rate(optimizer_gen)** be something...

tengolagan