{"id":15031119,"url":"https://github.com/tencent/facedetection-dsfd","last_synced_at":"2025-05-15T07:05:46.484Z","repository":{"id":38286821,"uuid":"162380526","full_name":"Tencent/FaceDetection-DSFD","owner":"Tencent","description":"腾讯优图高精度双分支人脸检测器","archived":false,"fork":false,"pushed_at":"2020-07-09T07:25:29.000Z","size":152200,"stargazers_count":2941,"open_issues_count":66,"forks_count":729,"subscribers_count":103,"default_branch":"master","last_synced_at":"2025-05-15T07:05:40.558Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Tencent.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"License.txt","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2018-12-19T03:59:06.000Z","updated_at":"2025-05-14T00:24:00.000Z","dependencies_parsed_at":"2022-08-09T02:30:52.114Z","dependency_job_id":null,"html_url":"https://github.com/Tencent/FaceDetection-DSFD","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Tencent%2FFaceDetection-DSFD","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Tencent%2FFaceDetection-DSFD/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Tencent%2FFaceDetection-DSFD/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Tencent%2FFaceDetection-DSFD/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Tencent","download_url":"https://codeload.github.com/Tencent/FaceDetection-DSFD/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":254292042,"owners_count":22046426,"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":[],"created_at":"2024-09-24T20:14:56.391Z","updated_at":"2025-05-15T07:05:41.400Z","avatar_url":"https://github.com/Tencent.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"\n\u003cimg src=\"imgs/DSFD_logo.PNG\" title=\"Logo\" width=\"300\" /\u003e \n\n## Update\n\n* 2019.04: Release pytorch-version DSFD inference code.\n* 2019.03: DSFD is accepted by CVPR2019.\n* 2018.10: Our DSFD ranks No.1 on [WIDER FACE](http://mmlab.ie.cuhk.edu.hk/projects/WIDERFace/WiderFace_Results.html) and [FDDB](http://vis-www.cs.umass.edu/fddb/results.html)\n\n\n## Introduction\n\u003cp align='center'\u003e\n  \u003cimg src='./imgs/dsfd_video.gif' width=1000'/\u003e\n\u003c/p\u003e\n\nIn this repo, we propose a novel face detection network, named DSFD, with superior performance over the state-of-the-art face detectors. You can use the code to evaluate our DSFD for face detection. \n\nFor more details, please refer to our paper [DSFD: Dual Shot Face Detector](https://arxiv.org/abs/1810.10220)! or poster [slide](./imgs/DSFD_CVPR2019_poster.pdf)!\n\n\u003cp align='center'\u003e\n\u003cimg src='./imgs/DSFD_framework.PNG' alt='DSFD Framework' width='1000px'\u003e\n\u003c/p\u003e\n\nOur DSFD face detector achieves state-of-the-art performance on [WIDER FACE](http://mmlab.ie.cuhk.edu.hk/projects/WIDERFace/WiderFace_Results.html) and [FDDB](http://vis-www.cs.umass.edu/fddb/results.html) benchmark.\n\n### WIDER FACE\n\u003cp align='center'\u003e\n\u003cimg src='./imgs/DSFD_widerface.PNG' alt='DSFD Widerface Performance' width='1000px'\u003e\n\u003c/p\u003e\n\n### FDDB\n\u003cp align='center'\u003e\n\u003cimg src='./imgs/DSFD_fddb.PNG' alt='DSFD FDDB Performance' width='1000px'\u003e\n\u003c/p\u003e\n\n## Requirements\n- Torch == 0.3.1\n- Torchvision == 0.2.1\n- Python == 3.6\n- NVIDIA GPU == Tesla P40 \n- Linux CUDA CuDNN\n\n## Getting Started\n\n### Installation\nClone the github repository. We will call the cloned directory as `$DSFD_ROOT`.\n```bash\ngit clone xxxxxx/FaceDetection-DSFD.git\ncd FaceDetection-DSFD\nexport CUDA_VISIBLE_DEVICES=0\n```\n\n\n### Evaluation\n1. Download the images of [WIDER FACE](http://mmlab.ie.cuhk.edu.hk/projects/WIDERFace/) and [FDDB](https://drive.google.com/open?id=17t4WULUDgZgiSy5kpCax4aooyPaz3GQH) to `$DSFD_ROOT/data/`.\n\n2. Download our DSFD model [微云](https://share.weiyun.com/567x0xQ) [google drive](https://drive.google.com/file/d/1WeXlNYsM6dMP3xQQELI-4gxhwKUQxc3-/view?usp=sharing) trained on WIDER FACE training set to `$DSFD_ROOT/weights/`.\n\n  \n3. Check out `./demo.py` on how to detect faces using the DSFD model and how to plot detection results.\n```\npython demo.py [--trained_model [TRAINED_MODEL]] [--img_root  [IMG_ROOT]] \n               [--save_folder [SAVE_FOLDER]] [--visual_threshold [VISUAL_THRESHOLD]] \n    --trained_model      Path to the saved model\n    --img_root           Path of test images\n    --save_folder        Path of output detection resutls\n    --visual_threshold   Confidence thresh\n```\n\n4. Evaluate the trained model via `./widerface_val.py` on WIDER FACE.\n```\npython widerface_val.py [--trained_model [TRAINED_MODEL]] [--save_folder [SAVE_FOLDER]] \n                         [--widerface_root [WIDERFACE_ROOT]]\n    --trained_model      Path to the saved model\n    --save_folder        Path of output widerface resutls\n    --widerface_root     Path of widerface dataset\n```\n\n5. Download the [eval_tool](http://mmlab.ie.cuhk.edu.hk/projects/WIDERFace/support/eval_script/eval_tools.zip) to show the WIDERFACE performance.\n\n6. Evaluate the trained model via `./fddb_test.py` on FDDB.\n```\npython widerface_test.py [--trained_model [TRAINED_MODEL]] [--split_dir [SPLIT_DIR]] \n                         [--data_dir [DATA_DIR]] [--det_dir [DET_DIR]]\n    --trained_model      Path of the saved model\n    --split_dir          Path of fddb folds\n    --data_dir           Path of fddb all images\n    --det_dir            Path to save fddb results\n```\n\n7. Download the [evaluation](http://vis-www.cs.umass.edu/fddb/evaluation.tgz) to show the FDDB performance.\n8. Lightweight DSFD is [here](https://github.com/lijiannuist/lightDSFD).\n\n## Qualitative Results\n\u003cp align='center'\u003e\n  \u003cimg src='./imgs/DSFD_demo1.PNG' width='1000'/\u003e\n\u003c/p\u003e\n\n\u003cp align='center'\u003e\n  \u003cimg src='./imgs/DSFD_demo2.PNG' width='1000'/\u003e\n\u003c/p\u003e\n\n\n### Citation\nIf you find DSFD useful in your research, please consider citing: \n```\n@inproceedings{li2018dsfd,\n  title={DSFD: Dual Shot Face Detector},\n  author={Li, Jian and Wang, Yabiao and Wang, Changan and Tai, Ying and Qian, Jianjun and Yang, Jian and Wang, Chengjie and Li, Jilin and Huang, Feiyue},\n  booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},\n  year={2019}\n}\n```\n## Contact\nFor any question, please file an issue or contact\n```\nJian Li: swordli@tencent.com\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftencent%2Ffacedetection-dsfd","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ftencent%2Ffacedetection-dsfd","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftencent%2Ffacedetection-dsfd/lists"}