{"id":25510656,"url":"https://github.com/guohengkai/region-ensemble-network","last_synced_at":"2025-04-10T15:54:09.571Z","repository":{"id":78296296,"uuid":"87610481","full_name":"guohengkai/region-ensemble-network","owner":"guohengkai","description":"Repository for Region Ensemble Network based Hand Pose Estimation","archived":false,"fork":false,"pushed_at":"2020-02-11T02:35:32.000Z","size":64860,"stargazers_count":114,"open_issues_count":2,"forks_count":36,"subscribers_count":10,"default_branch":"master","last_synced_at":"2025-03-24T13:44:04.830Z","etag":null,"topics":["convolutional-neural-networks","deep-learning","hand-pose-estimation","region-ensemble-network"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"gpl-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/guohengkai.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}},"created_at":"2017-04-08T05:28:23.000Z","updated_at":"2025-01-23T05:41:34.000Z","dependencies_parsed_at":"2023-02-28T04:31:20.819Z","dependency_job_id":null,"html_url":"https://github.com/guohengkai/region-ensemble-network","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/guohengkai%2Fregion-ensemble-network","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/guohengkai%2Fregion-ensemble-network/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/guohengkai%2Fregion-ensemble-network/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/guohengkai%2Fregion-ensemble-network/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/guohengkai","download_url":"https://codeload.github.com/guohengkai/region-ensemble-network/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248246067,"owners_count":21071625,"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":["convolutional-neural-networks","deep-learning","hand-pose-estimation","region-ensemble-network"],"created_at":"2025-02-19T09:52:10.433Z","updated_at":"2025-04-10T15:54:09.549Z","avatar_url":"https://github.com/guohengkai.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Towards Good Practices for Deep 3D Hand Pose Estimation\nBy Hengkai Guo (Updated on Aug 9, 2017)\n\n## Description\nThis is the project of work [Region Ensemble Network: Improving Convolutional Network for Hand Pose Estimation](https://arxiv.org/abs/1702.02447) and [Towards Good Practices for Deep 3D Hand Pose Estimation](https://arxiv.org/abs/1707.07248). This repository includes the prediction results for comparison, prediction codes and visualization codes. More details will be released in the future. Here are live results from Kinect 2 sensor using the model trained on ICVL:\n\n![result1.gif](demo/output_icvl_xinghao.gif) ![result2.gif](demo/output_icvl_hengkai.gif)\n\n## Results\nHere we provide the testing results of basic network (`results/dataset_basic.txt`) and region ensemble network (`results/dataset_ren_nx6x6.txt`) for [ICVL](http://www.iis.ee.ic.ac.uk/~dtang/hand.html) dataset, [NYU](http://cims.nyu.edu/~tompson/NYU_Hand_Pose_Dataset.htm) dataset and [MSRA](http://research.microsoft.com/en-us/um/people/yichenw/handpose/cvpr15_MSRAHandGestureDB.zip) dataset in our paper. Also we provide the testing labels (`labels/dataset_test_label.txt`), computed centers (`labels/dataset_center.txt`, which can be computed by `evaluation/get_centers.py`) and corresponding image names (`labels/dataset_test_list.txt`). Currently, the MSRA center computation is not available due to lack of loading function for images.\n\nFor results and labels, each line is corresponding to one image, which has J x 3 numbers indicating (x, y, z) of J joint locations. The (x, y) are in pixels and z is in mm.\n\n## Evaluation\nPlease use the Python script `evaluation/compute_error.py` for evaluation, which requires [numpy](http://www.numpy.org/) and [matplotlib](http://matplotlib.org/). Here is an example:\n``` bash\n$ python evaluation/compute_error.py icvl results/icvl_ren_9x6x6.txt\n```\n\n## Visualization\nPlease use the Python script `evaluation/show_result.py` for visualziation, which also requires [OpenCV](http://opencv.org/):\n``` bash\n$ python evaluation/show_result.py icvl your/path/to/ICVL/images/test/Depth --in_file=results/icvl_ren_4x6x6.txt\n```\nYou can see all the testing results on the images. Press 'q' to exit.\n\n## Prediction\nPlease use the Python script `evaluation/run_model.py` for prediction with predefined centers in `labels` directory:\n``` bash\n$ python evaluation/run_model.py icvl ren_4x6x6 your/path/to/output/file your/path/to/ICVL/images/test\n```\nThe script depends on [pyCaffe](https://github.com/BVLC/caffe). Please install the Caffe first.\n\n## Models\nThe caffe models can be downloaded at [BaiduYun](http://pan.baidu.com/s/1geFecSF), [GoogleDrive](https://drive.google.com/drive/folders/1wJERscsridu1GdStApBXXzh9G_dWTDne?usp=sharing) or [here](http://image.ee.tsinghua.edu.cn/data/icip2017-ren/models.zip). Please put them in the `models` directory. (For MSRA models, we only provide the one for fold 1 due to the limit of memory.)\n\n## Realsense Realtime Demo\nWe provide a realtime hand pose estimation demo using Intel Realsense device.\n\n### Using [pyrealsense](https://github.com/toinsson/pyrealsense)\nWhen you are using [pyrealsense](https://github.com/toinsson/pyrealsense) v1.x and v0.x, please use the Python script for demo:\n``` bash\n$ python demo/realsense_realtime_demo_pyrealsense_1.x.py\n```\nWhen you are using [pyrealsense](https://github.com/toinsson/pyrealsense) v2.0 and above, please use the Python script for demo:\n``` bash\n$ python demo/realsense_realtime_demo_pyrealsense_2.x.py\n```\n\n### Using [librealsense](https://github.com/IntelRealSense/librealsense)\nFirstly compile and install the [python wrapper](https://github.com/IntelRealSense/librealsense/tree/5285629b4ddb374f1). After everything is working properly, just run the following python script for demo:\n``` bash\n$ python demo/realsense_realtime_demo_librealsense2.py\n```\nNote that we just use a naive depth thresholding method to detect the hand. Therefore, the hand should be in the range of [0, 650mm] to run this demo.\nWe tested this realtime demo with an [Intel Realsense SR300](https://software.intel.com/en-us/realsense/sr300camera).\n\n## Citation\nPlease cite the paper in your publications if it helps your research:\n\n```\n@article{guo2017towards,\n  title={Towards Good Practices for Deep 3D Hand Pose Estimation},\n  author={Guo, Hengkai and Wang, Guijin and Chen, Xinghao and Zhang, Cairong},\n  journal={arXiv preprint arXiv:1707.07248},\n  year={2017}\n}\n@article{guo2017region,\n  title={Region Ensemble Network: Improving Convolutional Network for Hand Pose Estimation},\n  author={Guo, Hengkai and Wang, Guijin and Chen, Xinghao and Zhang, Cairong and Qiao, Fei and Yang, Huazhong},\n  journal={arXiv preprint arXiv:1702.02447},\n  year={2017}\n}\n```\n\n## License\nThis program is free software with GNU General Public License v2.\n\n## Feedback\nPlease email to `guohengkaighk@gmail.com` if you have any suggestions or questions.\n\n## History\nFeb 11, 2020: Update Google Drive link for models\n\nAug 9, 2017: Update papers\n\nJuly 23, 2017: Add script for center computing and results for newly paper\n\nMay 22, 2017: Intel Realsense realtime demo\n\nMay 15, 2017: More visualization and demos\n\nMay 9, 2017: Models and bugs fixed\n\nMay 6, 2017: Visualization and prediction codes\n\nApril 8, 2017: Evaluation codes\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fguohengkai%2Fregion-ensemble-network","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fguohengkai%2Fregion-ensemble-network","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fguohengkai%2Fregion-ensemble-network/lists"}