{"id":13627649,"url":"https://github.com/deepinx/iqiyi-vid-challenge","last_synced_at":"2025-04-17T00:32:08.384Z","repository":{"id":217074543,"uuid":"171210347","full_name":"deepinx/iqiyi-vid-challenge","owner":"deepinx","description":"Code for IQIYI-VID(IQIYI Video Person Identification) Challenge Implemented in Python and MXNet","archived":false,"fork":false,"pushed_at":"2019-04-22T01:55:39.000Z","size":606,"stargazers_count":47,"open_issues_count":1,"forks_count":13,"subscribers_count":4,"default_branch":"master","last_synced_at":"2024-11-08T18:45:03.081Z","etag":null,"topics":["face-alignment","face-detection","face-recognition","mxnet"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/deepinx.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"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}},"created_at":"2019-02-18T03:45:55.000Z","updated_at":"2023-08-31T12:09:55.000Z","dependencies_parsed_at":null,"dependency_job_id":"cb9da910-b5ac-4caf-980a-25f1299af1c9","html_url":"https://github.com/deepinx/iqiyi-vid-challenge","commit_stats":null,"previous_names":["deepinx/iqiyi-vid-challenge"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepinx%2Fiqiyi-vid-challenge","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepinx%2Fiqiyi-vid-challenge/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepinx%2Fiqiyi-vid-challenge/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepinx%2Fiqiyi-vid-challenge/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/deepinx","download_url":"https://codeload.github.com/deepinx/iqiyi-vid-challenge/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":249293004,"owners_count":21245666,"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":["face-alignment","face-detection","face-recognition","mxnet"],"created_at":"2024-08-01T22:00:36.771Z","updated_at":"2025-04-17T00:32:07.989Z","avatar_url":"https://github.com/deepinx.png","language":"Python","funding_links":[],"categories":["\u003ca name=\"Vision\"\u003e\u003c/a\u003e2. Vision"],"sub_categories":["2.5 Face Detection and Recognition"],"readme":"## Code for IQIYI-VID Challenge based on ESSH and Insightface\n\nRecently [www.iqiyi.com](http://www.iqiyi.com) released a great video person dataset called [IQIYI_VID](http://challenge.ai.iqiyi.com/detail?raceId=5afc36639689443e8f815f9e) and also launched a person search competition on it. It is a very large and real dataset worth trying to verify your face model accuracy precisely.\n\nThis repository contains the code for IQIYI-VID(IQIYI video person identification) Challenge. The methods are implemented in Python and MXNet. The Enhanced SSH (ESSH) from [enhanced-ssh-mxnet](https://github.com/deepinx/enhanced-ssh-mxnet) is applied for face detection and alignment. [Insightface](https://github.com/deepinsight/insightface) scheme is used for face recognition.\n\nPre-trained models can be downloaded on [BaiduCloud](https://pan.baidu.com/s/1vS_Uv9l5XZLeMwXPs_XzOA) or [GoogleDrive](https://drive.google.com/open?id=1_8-GcZbYNGvm_msyuwqDw4u9mGnHqSQP).\n\n## Environment\n\nThis repository has been tested under the following environment:\n\n-   Python 2.7 \n-   Ubuntu 18.04\n-   Mxnet-cu90 (==1.3.0)\n\n## Installation\n\n1.  Prepare the environment.\n\n2.  Clone the repository.\n    \n3.  Type  `make`  to build necessary cxx libs.\n\n4.  Download the pre-trained model and place it in *`./model/`*\n\n5.  Download the IQIYI-VID Datasets from [IQIYI_VID](http://challenge.ai.iqiyi.com/detail?raceId=5afc36639689443e8f815f9e) and unzip them to `data/iqiyi_vid` directory. \n\n## Usage\n\n1.  Detect faces on train+val dataset and test dataset respectively using ESSH model.  Model `model-r50-gg` is used to judge the quality of the detected faces.\n```\npython detect.py --model ./model/model-r50-gg/model,0 --output ./output/det_trainval --dataset ./data/iqiyi_vid --gpu 0 --stage trainval\npython detect.py --model ./model/model-r50-gg/model,0 --output ./output/det_test --dataset ./data/iqiyi_vid --gpu 0 --stage test\n```\n\n2. Extract features to the detected faces of train+val and test dataset respectively using `model-r100-gg` model.\n```\npython feature.py --model ./model/model-r100-gg/model,0 --input ./output/det_trainval --output ./output/feat_trainval  --gpu 0\npython feature.py --model ./model/model-r100-gg/model,0 --input ./output/det_test --output ./output/feat_test --gpu 0\n```\n3.  Re-save the extracted face features for training the MLP network.\n```\npython genfeat.py --inputs ./output/feat_trainval --output ./output/trainval\n```\n4. Train the MLP network for face ID recognition using train+val datasets.\n```\npython train_mlp.py --data ./output/trainval --prefix ./model/iqiyi --ckpt 1 --network r50 --lr 0.2 --per-batch-size 1024\n```\n5.  Predict face ID from features of the test dataset using the pre-trained MLP network.\n```\npython predict.py --model ./model/iqiyi,40 --gpu 0 --inputs ./output/feat_test --output ./output/pred_test\n```\n6. Run ``python submit.py`` to generate the final submissions for IQIYI-VID Challenge.\n\n## License\n\nMIT LICENSE\n\n\n## Reference\n\n```\n@article{deng2018arcface,\ntitle={ArcFace: Additive Angular Margin Loss for Deep Face Recognition},\nauthor={Deng, Jiankang and Guo, Jia and Niannan, Xue and Zafeiriou, Stefanos},\njournal={arXiv:1801.07698},\nyear={2018}\n}\n\n@inproceedings{Najibi2017SSH,\n  title={SSH: Single Stage Headless Face Detector},\n  author={Najibi, Mahyar and Samangouei, Pouya and Chellappa, Rama and Davis, Larry S.},\n  booktitle={IEEE International Conference on Computer Vision},\n  year={2017},\n}\n```\n\n## Acknowledgment\n\nThe code is adapted based on an intial fork from the [insightface](https://github.com/deepinsight/insightface) repository.\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdeepinx%2Fiqiyi-vid-challenge","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdeepinx%2Fiqiyi-vid-challenge","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdeepinx%2Fiqiyi-vid-challenge/lists"}