GitHub - yisol/IDM-VTON: [ECCV2024] IDM-VTON : Improving Diffusion Models for Authentic Virtual Try-on in the Wild

This is the official implementation of the paper "Improving Diffusion Models for Authentic Virtual Try-on in the Wild".

Star ⭐ us if you like it!

Requirements

git clone https://github.com/yisol/IDM-VTON.git
cd IDM-VTON

conda env create -f environment.yaml
conda activate idm

Data preparation

VITON-HD

You can download VITON-HD dataset from VITON-HD.

After download VITON-HD dataset, move vitonhd_test_tagged.json into the test folder, and move vitonhd_train_tagged.json into the train folder.

Structure of the Dataset directory should be as follows.


train
|-- image
|-- image-densepose
|-- agnostic-mask
|-- cloth
|-- vitonhd_train_tagged.json

test
|-- image
|-- image-densepose
|-- agnostic-mask
|-- cloth
|-- vitonhd_test_tagged.json

DressCode

You can download DressCode dataset from DressCode.

We provide pre-computed densepose images and captions for garments here.

We used detectron2 for obtaining densepose images, refer here for more details.

After download the DressCode dataset, place image-densepose directories and caption text files as follows.

DressCode
|-- dresses
    |-- images
    |-- image-densepose
    |-- dc_caption.txt
    |-- ...
|-- lower_body
    |-- images
    |-- image-densepose
    |-- dc_caption.txt
    |-- ...
|-- upper_body
    |-- images
    |-- image-densepose
    |-- dc_caption.txt
    |-- ...

Training

Preparation

Download pre-trained ip-adapter for sdxl(IP-Adapter/sdxl_models/ip-adapter-plus_sdxl_vit-h.bin) and image encoder(IP-Adapter/models/image_encoder) here.

git clone https://huggingface.co/h94/IP-Adapter

Move ip-adapter to ckpt/ip_adapter, and image encoder to ckpt/image_encoder.

Start training using python file with arguments,

accelerate launch train_xl.py \
    --gradient_checkpointing --use_8bit_adam \
    --output_dir=result --train_batch_size=6 \
    --data_dir=DATA_DIR

or, you can simply run with the script file.

Inference

VITON-HD

Inference using python file with arguments,

accelerate launch inference.py \
    --width 768 --height 1024 --num_inference_steps 30 \
    --output_dir "result" \
    --unpaired \
    --data_dir "DATA_DIR" \
    --seed 42 \
    --test_batch_size 2 \
    --guidance_scale 2.0

or, you can simply run with the script file.

DressCode

For DressCode dataset, put the category you want to generate images via category argument,

accelerate launch inference_dc.py \
    --width 768 --height 1024 --num_inference_steps 30 \
    --output_dir "result" \
    --unpaired \
    --data_dir "DATA_DIR" \
    --seed 42 
    --test_batch_size 2
    --guidance_scale 2.0
    --category "upper_body"

or, you can simply run with the script file.

Start a local gradio demo

Download checkpoints for human parsing here.

Place the checkpoints under the ckpt folder.

ckpt
|-- densepose
    |-- model_final_162be9.pkl
|-- humanparsing
    |-- parsing_atr.onnx
    |-- parsing_lip.onnx

|-- openpose
    |-- ckpts
        |-- body_pose_model.pth

Run the following command:

python gradio_demo/app.py

Acknowledgements

Thanks ZeroGPU for providing free GPU.

Thanks IP-Adapter for base codes.

Thanks OOTDiffusion and DCI-VTON for masking generation.

Thanks SCHP for human segmentation.

Thanks Densepose for human densepose.

Star History

Citation

@article{choi2024improving,
  title={Improving Diffusion Models for Authentic Virtual Try-on in the Wild},
  author={Choi, Yisol and Kwak, Sangkyung and Lee, Kyungmin and Choi, Hyungwon and Shin, Jinwoo},
  journal={arXiv preprint arXiv:2403.05139},
  year={2024}
}

License

The codes and checkpoints in this repository are under the CC BY-NC-SA 4.0 license.