{"id":13771944,"url":"https://github.com/bcmi/Accessory-Try-On-Dataset-STRAT","last_synced_at":"2025-05-11T04:31:02.579Z","repository":{"id":189336439,"uuid":"680495808","full_name":"bcmi/Accessory-Try-On-Dataset-STRAT","owner":"bcmi","description":"A virtual accessory try-on dataset which could be used for image composition ","archived":false,"fork":false,"pushed_at":"2024-02-14T14:11:05.000Z","size":278,"stargazers_count":21,"open_issues_count":0,"forks_count":0,"subscribers_count":5,"default_branch":"main","last_synced_at":"2024-11-17T07:34:32.383Z","etag":null,"topics":["image-composition","virtual-try-on"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/bcmi.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null}},"created_at":"2023-08-19T12:37:08.000Z","updated_at":"2024-10-12T08:18:47.000Z","dependencies_parsed_at":"2023-08-19T13:57:06.600Z","dependency_job_id":"856e6bc5-fe52-4f54-95ae-a01a89324701","html_url":"https://github.com/bcmi/Accessory-Try-On-Dataset-STRAT","commit_stats":null,"previous_names":["bcmi/accessory-try-on-dataset-strat"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bcmi%2FAccessory-Try-On-Dataset-STRAT","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bcmi%2FAccessory-Try-On-Dataset-STRAT/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bcmi%2FAccessory-Try-On-Dataset-STRAT/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/bcmi%2FAccessory-Try-On-Dataset-STRAT/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/bcmi","download_url":"https://codeload.github.com/bcmi/Accessory-Try-On-Dataset-STRAT/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":253518941,"owners_count":21921074,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["image-composition","virtual-try-on"],"created_at":"2024-08-03T17:00:57.915Z","updated_at":"2025-05-11T04:31:00.403Z","avatar_url":"https://github.com/bcmi.png","language":"Python","funding_links":[],"categories":["Datasets"],"sub_categories":[],"readme":"# Accessory-Try-On-Dataset-STRAT\n\nThis is the dataset used in the following research papers:\n\n* Spatial Transformation for Image Composition via Correspondence Learning [[arXiv]](https://arxiv.org/pdf/2207.02398.pdf)\n* Virtual Accessory Try-On via Keypoint Hallucination [[arXiv]](https://arxiv.org/pdf/2310.17131.pdf)\n\n## Dataset\nOur STRAT dataset contains three subdatasets: STRAT-glasses, STRAT-hat, and STRAT-tie, which correspond to \"glasses try-on\", \"hat try-on\", and \"tie try-on\" respectively. By treating accessory image (*resp.*, human face or portrait image) as foreground (*resp.*, background), we ask human annotators to paste the foreground on the background and adjust four vertices of foreground to generate a realistic composite image. After adjustment, the coordinates of four vertices are recorded. In each subdataset, the training set has **2000** pairs of foregrounds and backgrounds, while the test set has **1000** pairs of foregrounds and backgrounds. Some examples of foreground, background, and annotated composite image with composite foreground mask from three subdatasets are show in the figure below.\n\n\u003cimg src='example.jpg' align=\"center\" width=1024\u003e\n\n## Annotation\nIn the ``trans_params.json``, we record both the vertice coordinates and warpping parameters of the foreground for each pairwise sample. Specifically, ``init_vertex`` represents the four vertices of the accessory in foreground image, and ``final_vertex`` means the four vertices of the warped accessory in composite image. Given pairwise ``init_vertex`` and ``final_vertex``, the transformation matrix (``trans_matri``) can be derived by calling ``findHomography`` function of OpenCV library. To facilitate verification of the warpping parameters we provide or your own predicted parameters, we shared python scripts ``composite_by_opencv.py`` and ``composite_by_pytorch.py`` in this repository, which implement vertex transformation and composite image generation based on transformation matrix using opencv and pytorch, respectively. Considering that recent virtual try-on methods usually require semantic parsing, we also provide ground-truth semantic annotations in ``background_mask`` and ``background_mask_comp`` folders.     \n\n## Download\nThe STRAT dataset is provided in [[Baidu Cloud]](https://pan.baidu.com/s/13J7jyxexNausxwEvVgaAfg) (access code: bcmi) or [[Google Drive]](https://drive.google.com/file/d/1vHN1vvg2JQZKRfqrYfQjyuJNvgkP9-gY/view?usp=sharing).\n\n## Bibtex\n\nIf you use our dataset, please cite our paper using the following BibTeX  [[arxiv](https://arxiv.org/pdf/2207.02398.pdf)]:\n\n```\n@article{STRAT,\n  author       = {Bo Zhang and\n                  Yue Liu and\n                  Kaixin Lu and\n                  Li Niu and\n                  Liqing Zhang},\n  title        = {Spatial Transformation for Image Composition via Correspondence Learning},\n  journal      = {CoRR},\n  volume       = {abs/2207.02398},\n  year         = {2022},\n}\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbcmi%2FAccessory-Try-On-Dataset-STRAT","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fbcmi%2FAccessory-Try-On-Dataset-STRAT","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fbcmi%2FAccessory-Try-On-Dataset-STRAT/lists"}