{"id":23478176,"url":"https://github.com/mtli/bffl","last_synced_at":"2026-02-04T02:02:15.577Z","repository":{"id":141260930,"uuid":"120392406","full_name":"mtli/BFFL","owner":"mtli","description":"Code for Brute-Force Facial Landmark Analysis With A 140,000-Way Classifier :smiley:","archived":false,"fork":false,"pushed_at":"2018-02-11T01:54:39.000Z","size":3503,"stargazers_count":37,"open_issues_count":0,"forks_count":9,"subscribers_count":3,"default_branch":"master","last_synced_at":"2025-07-07T11:11:17.215Z","etag":null,"topics":["ai","computer-vision","facial-landmarks"],"latest_commit_sha":null,"homepage":"","language":"Matlab","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/mtli.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,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2018-02-06T02:46:27.000Z","updated_at":"2024-04-28T05:44:47.000Z","dependencies_parsed_at":null,"dependency_job_id":"4da4a6c3-103b-447b-b8af-307fbe25e08a","html_url":"https://github.com/mtli/BFFL","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/mtli/BFFL","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mtli%2FBFFL","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mtli%2FBFFL/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mtli%2FBFFL/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mtli%2FBFFL/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/mtli","download_url":"https://codeload.github.com/mtli/BFFL/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mtli%2FBFFL/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29064200,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-04T01:55:38.219Z","status":"online","status_checked_at":"2026-02-04T02:00:07.999Z","response_time":62,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"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":["ai","computer-vision","facial-landmarks"],"created_at":"2024-12-24T19:16:41.645Z","updated_at":"2026-02-04T02:02:15.571Z","avatar_url":"https://github.com/mtli.png","language":"Matlab","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Brute-Force Facial Landmark Analysis\n\n[[arXiv]](https://arxiv.org/abs/1802.01777)\n\n\u003cp align=\"center\"\u003e\u003cimg alt=\"Teaser\" src=\"doc/Teaser4.png\" width=\"500px\"\u003e\u003c/p\u003e\n\u003cp align=\"center\"\u003e\u003cimg alt=\"Visual\" src=\"doc/vis.png\" width=\"500px\"\u003e\u003c/p\u003e\n\n## Dependency\n* VLFeat\n* MatConvNet (tested with commit [d62881db](https://github.com/vlfeat/matconvnet/tree/d62881dbb587e4d5ed6750549b6a6b3f7559c84f))\n\n## Usage\n1. Download the [pre-trained model](https://drive.google.com/file/d/1oTRRsYseMnIXWiBR-OODuYm1I3NgdDvQ/view?usp=sharing) and extract to `models/`\n2. Run `Test.m`\n\n## Face detection\n\nThe detection for the example images are provided. However, to run on new images, a face detector is required. We recommend using [MTCNNv2](https://kpzhang93.github.io/MTCNN_face_detection_alignment/) due to its robustness and stability. Also, our detection refinement module is trained with MTCNNv2 using its default parameters.\n\nThe accepted format of the bounding box is [x y width height] (no need to round to integer), different from the output of the `detect_face` function in MTCNNv2. It can be transformed using the following code:\n\n```\nbbx(:, 3:4) = bbx(:, 3:4) - bbx(:, 1:2);\n```\n\n## Videos\n* [Temporal smoothing under complete occlusion (provided detection)](doc/HMM.mp4)\n* [Interactive conditional prediction - eye corner](doc/Interactive-Eye.mp4)\n* [Interactive conditional prediction - nose tip](doc/Interactive-Nose.mp4)\n\n\n## Citation\nIf you use this code for your research, please cite the paper:\n\n```\n@article{BFFL2018,\n  title={Brute-Force Facial Landmark Analysis With A 140,000-Way Classifier},\n  author={Li, Mengtian and Jeni, Laszlo and Ramanan, Deva},\n  journal={AAAI},\n  year={2018}\n}\n```","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmtli%2Fbffl","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmtli%2Fbffl","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmtli%2Fbffl/lists"}