Datasets:

liujx233
/

OneModelForAll_Preprocessed_Tryon_Data

license: other
task_categories:
  - image-to-image
  - text-to-image
tags:
  - virtual-try-on
  - virtual-try-off
  - fashion
  - preprocessing
pretty_name: OneModelForAll Preprocessed Try-on Supplementary Data

OneModelForAll Preprocessed Supplementary Data

This repository provides the preprocessed supplementary data used by One Model For All, a unified framework for virtual try-on, virtual try-off, and arbitrary-pose try-on.

The files are intended to be used together with the original VITON-HD and DeepFashion-MultiModal-Parts2Whole datasets. They include auxiliary data required by the OneModelForAll codebase, such as SMPL pose maps, head-region images, and the JSONL split files used for DeepFashion-MultiModal.

Note

This dataset card only describes the supplementary preprocessed files released for research use with OneModelForAll. Please follow the licenses and terms of the original datasets when using the data.

Acknowledgements

Thanks SCHP for human segmentation.

Thanks VITON-HD for providing the dataset.

Thanks DeepFashion-MultiModal and Parts2Whole for providing and processing the DeepFashion-Multimodal dataset used in this project.