嗨,按原作者默认参数,我复现了一把实验。 训练模型存放于百度网盘(链接: https://pan.baidu.com/s/1h6h9MYswltN4mcp5dfYycg 提取码: uwgg),供下载测试。 部分分析和拙见写在 Github: https://github.com/cinastanbean/cp-vton ,欢迎同行拍砖讨论。
Run the matlab codeconvert_data.munder the original data rootVITON/data, and get the new format. We use the json format for pose info as generated byOpenPose. Move these directories into our own datarootdata. You can get the processed data atGoogleDriveor by running: ...
^CP-VTON+: Clothing Shape and Texture Preserving Image-Based Virtual Try-On https://minar09.github.io/cpvtonplus/ ^LA-VITON: A Network for Looking-Attractive Virtual Try-On https://openaccess.thecvf.com/content_ICCVW_2019/papers/CVFAD/Lee_LA-VITON_A_Network_for_Looking-Attractive_Virtual...
嗨,按原作者默认参数,我复现了一把实验。训练模型存放于百度网盘(链接: https://pan.baidu.com/s/1h6h9MYswltN4mcp5dfYycg 提取码: uwgg),供下载测试。部分分析和拙见写在 Github: https://github.com/cinastanbean/cp-vton ,欢迎同行拍砖讨论。 $ tree checkpoints/ checkpoints/ ├── gmm_train_...