search

minar09 / cp-vton-plus

Official implementation for "CP-VTON+: Clothing Shape and Texture Preserving Image-Based Virtual Try-On", CVPRW 2020
https://minar09.github.io/cpvtonplus/
MIT License
346 stars 120 forks source link

Wrong output image while testing #57

Closed aeg12345 closed 3 years ago

aeg12345 commented 3 years ago

Greetings, When I try to generate images using the sample dataset but changing the image pairs I still get the same image outputs from the test dataset. I wanted to test this before trying out custom image.

pip env: home = /usr/local/bin include-system-site-packages = false version = 3.6.12

requirements.txt: torch==0.4.1

torchvision

torchvision==0.2.1 tensorboardX pillow numpy opencv-contrib-python

Sample Test pairs: 000001_0.jpg 019552_1.jpg 000010_0.jpg 004387_1.jpg 000020_0.jpg 019552_1.jpg

Result Dir outputs: 000001_0 000010_0 000020_0

Tryon outputs:

000001_0 000010_0 000020_0

Sample clothes: 019552_1 004387_1

The test pairs in the default test txt: 000001_0.jpg 001744_1.jpg

000010_0.jpg 004325_1.jpg 000020_0.jpg 002022_1.jpg

Test pair sample cloth images: 001744_1 004325_1 002022_1

I have made no changes to test py, the commands I use to run are:

python3.6 test.py --name GMM --stage GMM --batch-size 1 --datamode test --data_list test_pairs_new.txt --checkpoint checkpoints/GMM/gmm_final.pth

python3.6 test.py --name TOM --stage TOM --batch-size 1 --datamode test --data_list test_pairs_new.txt --checkpoint checkpoints/TOM/tom_final.pth

test_pairs_new.txt

aeg12345 commented 3 years ago

It was due to forgetting to move the generated GMM relevant folders to test folder.