search

sangyun884 / HR-VITON

Official PyTorch implementation for the paper High-Resolution Virtual Try-On with Misalignment and Occlusion-Handled Conditions (ECCV 2022).
825 stars 171 forks source link

image-parse-agnostic-v3.2 #68

Closed MosbehBarhoumi closed 1 year ago

MosbehBarhoumi commented 1 year ago

One of the authors stated that they utilized the get_parse_agnostic.py function from the GitHub repository https://github.com/sangyun884/HR-VITON/blob/main/get_parse_agnostic.py to obtain image-parse-agnostic-v3.2. However, I observed that the resulting data differed from the pre-processed data.

Blurjp commented 1 year ago

How did you figure this out?

SalehAhmad1 commented 5 months ago

What's the update on this issue? When we use HR-VTON's function to generate the "image-parse-agnostic-v3.2.", our data is grayscaled and not RGB like as in the zelando's preprocessed data