文件列表:

.idea (0, 2018-04-02)
.idea\MPPE.iml (554, 2018-04-02)
.idea\inspectionProfiles (0, 2018-04-02)
.idea\inspectionProfiles\Project_Default.xml (817, 2018-04-02)
.idea\misc.xml (406, 2018-04-02)
.idea\modules.xml (260, 2018-04-02)
.idea\other.xml (233, 2018-04-02)
ActionRecognition (0, 2018-04-02)
ActionRecognition\Actions (0, 2018-04-02)
ActionRecognition\Actions\boxing.m (1608389, 2018-04-02)
ActionRecognition\FisherVector.py (4334, 2018-04-02)
ActionRecognition\GMM.py (2706, 2018-04-02)
ActionRecognition\test_action_from_pose.py (3039, 2018-04-02)
ActionRecognition\tmp.py (1340, 2018-04-02)
ActionRecognition\train_action_from_pose.py (1848, 2018-04-02)
PosturalRecognition (0, 2018-04-02)
PosturalRecognition\models (0, 2018-04-02)
PosturalRecognition\models\FaceNet.py (8609, 2018-04-02)
PosturalRecognition\models\HandNet.py (8602, 2018-04-02)
PosturalRecognition\models\PoseNet.py (15374, 2018-04-02)
PosturalRecognition\models\__init__.py (0, 2018-04-02)
PosturalRecognition\models\haarcascade_frontalface_alt.xml (674605, 2018-04-02)
PosturalRecognition\plot_train_log.py (690, 2018-04-02)
PosturalRecognition\test (0, 2018-04-02)
PosturalRecognition\test\VideoCapture.py (4093, 2018-04-02)
PosturalRecognition\test\__init__.py (0, 2018-04-02)
PosturalRecognition\test\camera_face.py (2315, 2018-04-02)
PosturalRecognition\test\camera_pose.py (1770, 2018-04-02)
PosturalRecognition\test\get_person_pose_array.py (7036, 2018-04-02)
PosturalRecognition\test\save_npy.py (2937, 2018-04-02)
PosturalRecognition\train (0, 2018-04-02)
PosturalRecognition\train\__init__.py (0, 2018-04-02)
PosturalRecognition\train\coco_data_loader.py (27588, 2018-04-02)
PosturalRecognition\train\data_load_test.py (484, 2018-04-02)
PosturalRecognition\train\entity.py (4050, 2018-04-02)
PosturalRecognition\train\face_detector.py (5604, 2018-04-02)
PosturalRecognition\train\gen_ignore_mask.py (5044, 2018-04-02)
PosturalRecognition\train\hand_detector.py (5262, 2018-04-02)
PosturalRecognition\train\pose_detector.py (23648, 2018-04-02)
... ...

# 基于Intel Chainer 和姿势检测的动作识别 ## 环境要求 - Python 3.0+ - Intel Chainer - NumPy - Six - Swig - Matplotlib - OpenCV - COCO API ## 目录结构 - ActionRecognition - Actions - data - KTH - dpt - jabbing - datasets - coco - KTH - UCF101 - img - PosturalRecognition - models - test - train - result ## 安装 COCO API ``` git clone https://github.com/cocodataset/cocoapi.git cd cocoapi/PythonAPI/ make python setup.py install cd ../../ ``` 如果在windows系统上安装,必须将VS中的VC++添加到系统PATH,否则会出现找不到相应文件 ## 获取COCO数据集 ``` mkdir datasets cd datasets mkdir coco cd coco wget http://images.cocodataset.org/zips/train2017.zip wget http://images.cocodataset.org/zips/val2017.zip wget http://images.cocodataset.org/annotations/annotations_trainval2017.zip unzip train2017.zip unzip val2017.zip unzip annotations_trainval2017.zip rm -f train2017.zip rm -f val2017.zip rm -f annotations_trainval2017.zip ``` 在Intel devcloud 上应该检查解压完全完成后才可以删除压缩包 若需要训练,必须先做数据预处理,在预处理之后执行文件即可 ``` python3 train_coco_pose_estimation.py ``` ## 测试 ``` python PosturalRecognition/test/get_person_pose_array.py ``` 若要测试自己的图片,应当修改main,若需要绘制其他效果可以取消部分注释的代码   ## 动作识别 ### 训练 ``` python ActionRecognition/train_action_from_pose.py ``` ### 测试 通过get_person_pose_array获取信息,然后运行 ``` python ActionRecognition/test_action_from_pose.py ``` 其中,`person02_boxing_d2_uncomp` 和 `person05_walking_d1_uncomp` 需要预先通过 `PosturalRecognition/test/VideoCapture.py` 进行预处理 最后输出动作发生的概率

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：