python main.py
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
/home/sarthak/anaconda3/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
Traceback (most recent call last):
File "main.py", line 6, in
from utils import choose_run_mode, load_pretrain_model, set_video_writer
File "/home/sarthak/Major Project/Online-Realtime-Action-Recognition-based-on-OpenPose/utils.py", line 6, in
from Pose.pose_visualizer import TfPoseVisualizer
File "/home/sarthak/Major Project/Online-Realtime-Action-Recognition-based-on-OpenPose/Pose/pose_visualizer.py", line 6, in
from .pose_estimator import estimate
File "/home/sarthak/Major Project/Online-Realtime-Action-Recognition-based-on-OpenPose/Pose/pose_estimator.py", line 21, in
'coord1', 'coord2', 'score1', 'score2'], verbose=False)
TypeError: namedtuple() got an unexpected keyword argument 'verbose'

I have trained the model using 4 classes(sit,stand,walk and fall down).For each class,I have created about 3000 samples and converted it to csv file and run the training for 1000 epochs.While testing,the actions are mismatched.i.e.,sit is shown as stand/walk/fall down.I have seen your results in this page and found that the actions are accurately detected.
Can you help me how you gain those results?

I want to ask you what does it mean to have 36 decimals in a CSV file?What is the meaning of the 18 key points?Some of the pose estimates that we've seen before are outputs of coordinates.Thank you.

I am giving my own video file but error appears that "Float is required" in main.py set_video_writer function. Also I am uploading scree

nshot for your reference.

读网络rtmp或本地文件，（1）识别的动作只有operate，（2）运行一段时间都会卡死
请问你有遇到这样的情况吗，怎么解决的？

Thanks for your good job, II want to konw tensorflow，keras and other tools version？Thank you~

您好，我想看看您这篇代码对应的论文，您能发一下链接吗，谢谢

Hello,
Thanks for the great work on this repo!
I am trying to run the code on my Windows machine with Anaconda Python but I have 2 issues:

1. I can't download the VGG tf-model from the baidu website link nor does it download when using "./download" command! so I went to the mediafire links to download the VGG tf-model.
2. now when I run "main.py" I get this error:

Traceback (most recent call last):
  File "main.py", line 16, in <module>
    action_classifier = load_action_premodel('Action/framewise_recognition.h5')
  File "C:\Users\malabdou\Documents\Python\PyProj\Posture\10Online-Realtime-Action-Recognition-based-on-OpenPose-master\Action\recognizer.py", line 99, in load_action_premodel
    return load_model(model)
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\keras\engine\saving.py", line 260, in load_model
    model = model_from_config(model_config, custom_objects=custom_objects)
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\keras\engine\saving.py", line 334, in model_from_config
    return deserialize(config, custom_objects=custom_objects)
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\keras\layers\__init__.py", line 55, in deserialize
    printable_module_name='layer')
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\keras\utils\generic_utils.py", line 145, in deserialize_keras_object
    list(custom_objects.items())))
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\keras\engine\sequential.py", line 293, in from_config
    model.add(layer)
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\keras\engine\sequential.py", line 185, in add
    output_tensor = layer(self.outputs[0])
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\keras\engine\base_layer.py", line 457, in __call__
    output = self.call(inputs, **kwargs)
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\keras\layers\core.py", line 878, in call
    output = self.activation(output)
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\keras\activations.py", line 29, in softmax
    return K.softmax(x)
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\keras\backend\tensorflow_backend.py", line 3154, in softmax
    return tf.nn.softmax(x, axis=axis)
TypeError: softmax() got an unexpected keyword argument 'axis'

Any way to solve this please?

Awesome project!
请问作者近期有没有用RNN/LSTM加入时序来做呢？

Hello, when I run the model to test, the FPS is very low, and the calculation speed of the running model is very slow. Can there be a way to speed up the operation?

Could you give the version of tensorflow, I got an error about the "NodeDef mentions attr "dilations" not in Op".

Hi, Thanks for the great implementation.
However, it is super slow on my laptop (FPS 0.5 ). I dont't have NVIDIA GPU, so I was wondering if this is the issue ?
I appreciate if you help me to fix it.
My system configuration:
OS: Ubunutu 18.04
CPU: Intel(R) Core(TM) i7-8550U CPU @ 1.80GHz
GPU: Intel® UHD Graphics 620
RAM: 16GB

thanks

您好，请问一下：
你们的训练集是怎么得到的呢？我们通过阅读你文章中关于训练的那部分，然后训练数据集的出来的数据，跟你的“data.csv”相比，我们数据中的“0”比较多，然后我们测了下我们的精度没有你们那个高。
你们的大概是97%，而我们只有80%。请问一下你们当时是怎样提高精度的呢？使用了什么数据集进行训练呢？
非常感谢！！！

Hi, I have trained the model with only "sit" action with 2500 samples. I have an issue that model detects the other actions also as sit. Is there any solution to detect other actions as "N/A" without making 2 class label in training ("sit" and "N/A"). Find the results( only one action has been trained, where results shows every other action as "sit") below.

您好，我在测试的时候发现，在每一秒都会输出一个类别，这个类别有时候根本不对，怎么样才能让准确率大于90或者更多时，才输出类别的名称呢？比如我没有训练人正常走路的类别，在视频中出现人正常走路时，就会被识别成别的种类，可以让他在遇到没有训练过的动作时，不输出类别名称么？

作者你好，首先非常感谢代码的分享！我有个问题想请教一下：我只在姿态估计部分(pose_visualizer.py)看到了创建tf会话的语句（但没找到结束会话的语句？），而且为什么在行为识别部分不需要再创建keras会话呢？非常期待收到您的回复，谢谢！

Hello ,
Please when i changed the enum of action in the units i have this erreur : Error when checking target: expected dense_4 to have shape (14,) but got array with shape (4,)

Please can you help me

你好，在运行train.py时程序出现了这个问题
ValueError: The first layer in a Sequential model must get an input_shape or batch_input_shape argument.
是在keras的model.py中的问题，是不是train.py中漏了什么语句呢？

I can't find VGG_Origin model file in the repository . Instead of .pb file I can only see a shell file there. Can someone please tell me how to load the model.
The code works with mobilenet model though.

I have glanced at the description of OpenPose.The function of 2D real-time multi-person keypoint detection have been done by OpenPose.Why do you detect persons and track them to recognitize pose of one person?Why not detect the skeletons of multiple persons using openpose directly?It is skeletons that pose recognition based on，isn't it?Thank you for answering my questions.
(ps:I am a newcomer of picture processing.So maybe my ideas are wrong.)

你好，我想问一下你们的数据是怎样采集的，数据采集的要求是什么，是不是画面中必须是单人？而且用excle处理数据具体是怎么做的呢？感觉很麻烦？

Hi,
quoted from READme.md "USAGE" section:
Download the openpose VGG tf-model with command line ./download.sh(/Pose/graph_models/VGG_origin) or fork here, and place it under the corresponding folder;

I think both the fork link and Mediafire link is written in bash file are down. Any other alternative, please?

Thanks for your cool work, II want to ask how to use this method on other action classifications, how to prepare the data set and train them?

Would you let me know the minimum hardware requirements or your test system specifications?

_____ it doesn't give any errors but a bunch of warnings and stops at the end for a sec and I can see my webcam light come for a sec then just returns back to prev command line.
___________________________________>>>>>>

D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master>python main.py
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_quint16 = np.dtype([("quint16", np.uint16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_qint32 = np.dtype([("qint32", np.int32, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\sklearn\utils\linear_assignment.py:22: FutureWarning: The linear_assignment module is deprecated in 0.21 and will be removed from 0.23. Use scipy.optimize.linear_sum_assignment instead.
FutureWarning)
Using TensorFlow backend.
WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:71: The name tf.gfile.GFile is deprecated. Please use tf.io.gfile.GFile instead.

WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:72: The name tf.GraphDef is deprecated. Please use tf.compat.v1.GraphDef instead.

WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:75: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.

WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:77: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead.

2020-03-19 15:45:20.504278: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
Traceback (most recent call last):
File "main.py", line 15, in
estimator = load_pretrain_model('VGG_origin')
File "D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\utils.py", line 46, in load_pretrain_model
raise Exception('Graph file doesn't exist, path=%s' % graph_path)
Exception: Graph file doesn't exist, path=D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Pose\graph_models\VGG_origin\graph_opt.pb

D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master>python main.py
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_quint16 = np.dtype([("quint16", np.uint16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_qint32 = np.dtype([("qint32", np.int32, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\sklearn\utils\linear_assignment.py:22: FutureWarning: The linear_assignment module is deprecated in 0.21 and will be removed from 0.23. Use scipy.optimize.linear_sum_assignment instead.
FutureWarning)
Using TensorFlow backend.
WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:71: The name tf.gfile.GFile is deprecated. Please use tf.io.gfile.GFile instead.

WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:72: The name tf.GraphDef is deprecated. Please use tf.compat.v1.GraphDef instead.

WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:75: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.

WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:77: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead.

2020-03-19 15:46:02.547502: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
WARNING:tensorflow:From C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\keras\backend\tensorflow_backend.py:422: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.

D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master>python main.py
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint16 = np.dtype([("quint16", np.uint16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint32 = np.dtype([("qint32", np.int32, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\framework\dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint8 = np.dtype([("qint8", np.int8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_quint8 = np.dtype([("quint8", np.uint8, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
_np_qint16 = np.dtype([("qint16", np.int16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_quint16 = np.dtype([("quint16", np.uint16, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_qint32 = np.dtype([("qint32", np.int32, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
np_resource = np.dtype([("resource", np.ubyte, 1)])
C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\sklearn\utils\linear_assignment.py:22: FutureWarning: The linear_assignment module is deprecated in 0.21 and will be removed from 0.23. Use scipy.optimize.linear_sum_assignment instead.
FutureWarning)
Using TensorFlow backend.
WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:71: The name tf.gfile.GFile is deprecated. Please use tf.io.gfile.GFile instead.

WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:72: The name tf.GraphDef is deprecated. Please use tf.compat.v1.GraphDef instead.

WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:75: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.

WARNING:tensorflow:From D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Tracking\generate_dets.py:77: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead.

2020-03-19 15:46:40.312092: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
WARNING:tensorflow:From C:\Users\magir\AppData\Local\Programs\Python\Python35\lib\site-packages\keras\backend\tensorflow_backend.py:422: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.

D:\Online-Realtime-Action-Recognition-based-on-OpenPose-master\Online-Realtime-Action-Recognition-based-on-OpenPose-master>

I am trying to run the original code but I do not find the pre-trained weights on the given link .can any guide me how to download the pre-trained weights

Dears, as shown on the training, how did you get or create the dataset.
best regards

I have two questions, I hope you can answer
How to use the webcam, no matter how to change the parameters, it is the default camera.
The next question, my ubuntu computer can't start the program, even if I set config,
Still can't start cudnn

run main.py,an error has occurred
Using TensorFlow backend.
Traceback (most recent call last):
File "F:/action recognition demo/OPEN-POSE/Online-Realtime-Action-Recognition-based-on-OpenPose/main.py", line 8, in
from Action.recognizer import load_action_premodel, framewise_recognize
File "F:\action recognition demo\OPEN-POSE\Online-Realtime-Action-Recognition-based-on-OpenPose\Action\recognizer.py", line 10, in
from keras.models import load_model
File "C:\Users\Administrator\Anaconda3\lib\site-packages\keras_init_.py", line 3, in
from . import utils
File "C:\Users\Administrator\Anaconda3\lib\site-packages\keras\utils_init_.py", line 6, in
from . import conv_utils
File "C:\Users\Administrator\Anaconda3\lib\site-packages\keras\utils\conv_utils.py", line 9, in
from .. import backend as K
File "C:\Users\Administrator\Anaconda3\lib\site-packages\keras\backend_init_.py", line 1, in
from .load_backend import epsilon
File "C:\Users\Administrator\Anaconda3\lib\site-packages\keras\backend\load_backend.py", line 90, in
from .tensorflow_backend import *
File "C:\Users\Administrator\Anaconda3\lib\site-packages\keras\backend\tensorflow_backend.py", line 13, in
from tensorflow.python.keras.utils import tf_utils
ImportError: cannot import name 'tf_utils'

You mean that using LSTM training to identify joint points will be better for classifying actions.

Hi，
我学习了一下代码，请教一个问题。代码中行为分类的特征向量是
record_joints_norm += [round(center_x/1280, 2), round(center_y/720, 2)]
这个归一化向量是不是对位置敏感呢？比如同样的姿势（Action），目标站在画面不同位置，特征向量是不同的吧？这样跟我们的目标是不是不太相同，因为我们显然希望不管目标站在画面何处，只要动作相同，分类都是相同的。

谢谢您的贡献和时间：）

Dear author
I have read your instructions about training with own dataset. I also have seen the data you used in your project. The question is the last column class in data.csv is manually generated by yourself?

I would like to ask how to change the name of the displayed action to Chinese.Thank you.

hello,how did you get the coordinates of these key points and normalize these numbers

Thank you for the python code. Please can you help me, how can I prepare and save my videos data to .txt
Thank you.

I have tried this model for my own data set (i have given sit action video and converted to csv file ).
Csv file contains 1000 records. Now in action_enum.py I have only sit action . In train.py file i have changed the encoder_Y = [0]*744 + [1]*722 + [2]*815 + [3]*1008 to
encoder_Y = [0]*1000 ( only for sit action ), After changing in train.py file,I got this error.

Is it possible to use the body_25 model from the openpose repo?

I want to train a model to recognize the stand walk operation fall_down four actions in the video. Would I open train.py and then modify the following three sentences:

1. model. save ('framewise_recognition. h5') - model. save ('framewise_recognition. h5')

2. model. fit (X_train, Y_train, batch_size=32, epochs=20, verbose=1, validation_data= (X_test, Y_test))-

Model. fit (X_train, Y_train, batch_size=32, epochs=20, verbose=1, validation_data= (X_test, Y_test), callbacks= [his])

3. raw_data = pd. read_csv ('data_with_scene. csv', header = 0)-

Raw_data = pd. read_csv ('data_under_scene. csv', header = 0)

Is that the case?

Thank you for your project
 What if I want to change the model to RNN (LSTM)？

When I downloaded your file and decompressed it, I wanted to retrain the model, open train.py, change the phrase # model.save ('zyhframewise_recognition.h5') to model.save ('zyhframewise_recognition.h5'), and run train.py.
Error occurred:

C:\ProgramData\Anaconda3\pythonw.exe H:/Online-Realtime-Action-Recognition-based-on-OpenPose-master/Online-Realtime-Action-Recognition-based-on-OpenPose-master/Action/training/train.py
Using TensorFlow backend.
Train on 2960 samples, validate on 329 samples
Epoch 1/20
2019-03-20 17:55:58.365067: I T:\src\github\tensorflow\tensorflow\core\platform\cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2

32/2960 [..............................] - ETA: 2:03 - loss: 1.2098 - acc: 0.3438
768/2960 [======>.......................] - ETA: 3s - loss: 1.1682 - acc: 0.4544
1344/2960 [============>.................] - ETA: 1s - loss: 1.0412 - acc: 0.5417
1952/2960 [==================>...........] - ETA: 0s - loss: 0.9448 - acc: 0.6071
2592/2960 [=========================>....] - ETA: 0s - loss: 0.8621 - acc: 0.6655
2912/2960 [============================>.] - ETA: 0s - loss: 0.8351 - acc: 0.6854
2960/2960 [==============================] - 2s 575us/step - loss: 0.8304 - acc: 0.6892 - val_loss: 0.5008 - val_acc: 0.8906
Epoch 2/20

32/2960 [..............................] - ETA: 0s - loss: 0.5474 - acc: 0.9375
544/2960 [====>.........................] - ETA: 0s - loss: 0.5454 - acc: 0.8750
928/2960 [========>.....................] - ETA: 0s - loss: 0.5366 - acc: 0.8804
1280/2960 [===========>..................] - ETA: 0s - loss: 0.5164 - acc: 0.8867
1760/2960 [================>.............] - ETA: 0s - loss: 0.4962 - acc: 0.8949
2144/2960 [====================>.........] - ETA: 0s - loss: 0.4804 - acc: 0.9002
2592/2960 [=========================>....] - ETA: 0s - loss: 0.4697 - acc: 0.9028
2960/2960 [==============================] - 0s 133us/step - loss: 0.4642 - acc: 0.9037 - val_loss: 0.3654 - val_acc: 0.9210
Epoch 3/20

32/2960 [..............................] - ETA: 0s - loss: 0.4151 - acc: 0.8750
544/2960 [====>.........................] - ETA: 0s - loss: 0.3649 - acc: 0.9412
1024/2960 [=========>....................] - ETA: 0s - loss: 0.3741 - acc: 0.9316
1312/2960 [============>.................] - ETA: 0s - loss: 0.3776 - acc: 0.9291
1632/2960 [===============>..............] - ETA: 0s - loss: 0.3729 - acc: 0.9295
1920/2960 [==================>...........] - ETA: 0s - loss: 0.3680 - acc: 0.9297
2336/2960 [======================>.......] - ETA: 0s - loss: 0.3640 - acc: 0.9315
2720/2960 [==========================>...] - ETA: 0s - loss: 0.3657 - acc: 0.9298
2960/2960 [==============================] - 0s 145us/step - loss: 0.3641 - acc: 0.9304 - val_loss: 0.2952 - val_acc: 0.9544
Epoch 4/20

32/2960 [..............................] - ETA: 0s - loss: 0.3248 - acc: 0.9062
480/2960 [===>..........................] - ETA: 0s - loss: 0.3031 - acc: 0.9521
960/2960 [========>.....................] - ETA: 0s - loss: 0.2900 - acc: 0.9573
1376/2960 [============>.................] - ETA: 0s - loss: 0.2961 - acc: 0.9520
1696/2960 [================>.............] - ETA: 0s - loss: 0.2931 - acc: 0.9522
2176/2960 [=====================>........] - ETA: 0s - loss: 0.2929 - acc: 0.9508
2816/2960 [===========================>..] - ETA: 0s - loss: 0.2826 - acc: 0.9528
2960/2960 [==============================] - 0s 111us/step - loss: 0.2899 - acc: 0.9490 - val_loss: 0.2277 - val_acc: 0.9696
Epoch 5/20

32/2960 [..............................] - ETA: 0s - loss: 0.1878 - acc: 0.9688
576/2960 [====>.........................] - ETA: 0s - loss: 0.2421 - acc: 0.9635
1056/2960 [=========>....................] - ETA: 0s - loss: 0.2417 - acc: 0.9602
1440/2960 [=============>................] - ETA: 0s - loss: 0.2466 - acc: 0.9563
2016/2960 [===================>..........] - ETA: 0s - loss: 0.2360 - acc: 0.9608
2656/2960 [=========================>....] - ETA: 0s - loss: 0.2243 - acc: 0.9646
2960/2960 [==============================] - 0s 104us/step - loss: 0.2252 - acc: 0.9659 - val_loss: 0.1825 - val_acc: 0.9696
Epoch 6/20

32/2960 [..............................] - ETA: 0s - loss: 0.1284 - acc: 1.0000
608/2960 [=====>........................] - ETA: 0s - loss: 0.1948 - acc: 0.9803
1216/2960 [===========>..................] - ETA: 0s - loss: 0.1889 - acc: 0.9811
1888/2960 [==================>...........] - ETA: 0s - loss: 0.1986 - acc: 0.9740
2528/2960 [========================>.....] - ETA: 0s - loss: 0.1925 - acc: 0.9735
2960/2960 [==============================] - 0s 91us/step - loss: 0.1917 - acc: 0.9713 - val_loss: 0.1518 - val_acc: 0.9696
Epoch 7/20

32/2960 [..............................] - ETA: 0s - loss: 0.1956 - acc: 0.9688
608/2960 [=====>........................] - ETA: 0s - loss: 0.1691 - acc: 0.9803
1216/2960 [===========>..................] - ETA: 0s - loss: 0.1670 - acc: 0.9786
1824/2960 [=================>............] - ETA: 0s - loss: 0.1634 - acc: 0.9753
2208/2960 [=====================>........] - ETA: 0s - loss: 0.1652 - acc: 0.9760
2752/2960 [==========================>...] - ETA: 0s - loss: 0.1648 - acc: 0.9760
2960/2960 [==============================] - 0s 102us/step - loss: 0.1629 - acc: 0.9770 - val_loss: 0.1257 - val_acc: 0.9726
Epoch 8/20

32/2960 [..............................] - ETA: 0s - loss: 0.0994 - acc: 1.0000
640/2960 [=====>........................] - ETA: 0s - loss: 0.1458 - acc: 0.9859
1120/2960 [==========>...................] - ETA: 0s - loss: 0.1404 - acc: 0.9866
1536/2960 [==============>...............] - ETA: 0s - loss: 0.1369 - acc: 0.9857
1792/2960 [=================>............] - ETA: 0s - loss: 0.1376 - acc: 0.9833
2208/2960 [=====================>........] - ETA: 0s - loss: 0.1381 - acc: 0.9823
2688/2960 [==========================>...] - ETA: 0s - loss: 0.1361 - acc: 0.9829
2960/2960 [==============================] - 0s 127us/step - loss: 0.1367 - acc: 0.9824 - val_loss: 0.1088 - val_acc: 0.9787
Epoch 9/20

32/2960 [..............................] - ETA: 0s - loss: 0.0539 - acc: 1.0000
480/2960 [===>..........................] - ETA: 0s - loss: 0.1378 - acc: 0.9812
768/2960 [======>.......................] - ETA: 0s - loss: 0.1311 - acc: 0.9857
1152/2960 [==========>...................] - ETA: 0s - loss: 0.1270 - acc: 0.9844
1856/2960 [=================>............] - ETA: 0s - loss: 0.1259 - acc: 0.9833
2336/2960 [======================>.......] - ETA: 0s - loss: 0.1287 - acc: 0.9824
2624/2960 [=========================>....] - ETA: 0s - loss: 0.1289 - acc: 0.9832
2960/2960 [==============================] - 0s 136us/step - loss: 0.1302 - acc: 0.9824 - val_loss: 0.0963 - val_acc: 0.9757
Epoch 10/20

32/2960 [..............................] - ETA: 0s - loss: 0.2233 - acc: 0.9062
544/2960 [====>.........................] - ETA: 0s - loss: 0.1215 - acc: 0.9688
960/2960 [========>.....................] - ETA: 0s - loss: 0.1116 - acc: 0.9750
1376/2960 [============>.................] - ETA: 0s - loss: 0.1100 - acc: 0.9767
1856/2960 [=================>............] - ETA: 0s - loss: 0.1103 - acc: 0.9784
2400/2960 [=======================>......] - ETA: 0s - loss: 0.1096 - acc: 0.9788
2960/2960 [==============================] - 0s 114us/step - loss: 0.1077 - acc: 0.9807 - val_loss: 0.0857 - val_acc: 0.9818
Epoch 11/20

32/2960 [..............................] - ETA: 0s - loss: 0.1307 - acc: 0.9688
448/2960 [===>..........................] - ETA: 0s - loss: 0.1064 - acc: 0.9754
896/2960 [========>.....................] - ETA: 0s - loss: 0.1025 - acc: 0.9799
1472/2960 [=============>................] - ETA: 0s - loss: 0.0977 - acc: 0.9837
2048/2960 [===================>..........] - ETA: 0s - loss: 0.1011 - acc: 0.9829
2496/2960 [========================>.....] - ETA: 0s - loss: 0.0992 - acc: 0.9844
2816/2960 [===========================>..] - ETA: 0s - loss: 0.0995 - acc: 0.9847
2960/2960 [==============================] - 0s 116us/step - loss: 0.0990 - acc: 0.9851 - val_loss: 0.0792 - val_acc: 0.9878
Epoch 12/20

32/2960 [..............................] - ETA: 0s - loss: 0.0712 - acc: 1.0000
608/2960 [=====>........................] - ETA: 0s - loss: 0.1006 - acc: 0.9836
1024/2960 [=========>....................] - ETA: 0s - loss: 0.0955 - acc: 0.9854
1472/2960 [=============>................] - ETA: 0s - loss: 0.0891 - acc: 0.9878
2016/2960 [===================>..........] - ETA: 0s - loss: 0.0876 - acc: 0.9876
2592/2960 [=========================>....] - ETA: 0s - loss: 0.0916 - acc: 0.9857
2960/2960 [==============================] - 0s 103us/step - loss: 0.0911 - acc: 0.9851 - val_loss: 0.0724 - val_acc: 0.9848
Epoch 13/20

32/2960 [..............................] - ETA: 0s - loss: 0.0867 - acc: 1.0000
384/2960 [==>...........................] - ETA: 0s - loss: 0.0909 - acc: 0.9740
576/2960 [====>.........................] - ETA: 0s - loss: 0.0871 - acc: 0.9792
928/2960 [========>.....................] - ETA: 0s - loss: 0.0879 - acc: 0.9817
1504/2960 [==============>...............] - ETA: 0s - loss: 0.0888 - acc: 0.9827
2080/2960 [====================>.........] - ETA: 0s - loss: 0.0836 - acc: 0.9856
2336/2960 [======================>.......] - ETA: 0s - loss: 0.0826 - acc: 0.9863
2880/2960 [============================>.] - ETA: 0s - loss: 0.0807 - acc: 0.9875
2960/2960 [==============================] - 0s 134us/step - loss: 0.0816 - acc: 0.9872 - val_loss: 0.0699 - val_acc: 0.9848
Epoch 14/20

32/2960 [..............................] - ETA: 0s - loss: 0.0764 - acc: 0.9688
608/2960 [=====>........................] - ETA: 0s - loss: 0.0739 - acc: 0.9885
1184/2960 [===========>..................] - ETA: 0s - loss: 0.0807 - acc: 0.9865
1600/2960 [===============>..............] - ETA: 0s - loss: 0.0780 - acc: 0.9894
2016/2960 [===================>..........] - ETA: 0s - loss: 0.0736 - acc: 0.9916
2592/2960 [=========================>....] - ETA: 0s - loss: 0.0763 - acc: 0.9900
2960/2960 [==============================] - 0s 100us/step - loss: 0.0763 - acc: 0.9902 - val_loss: 0.0651 - val_acc: 0.9848
Epoch 15/20

32/2960 [..............................] - ETA: 0s - loss: 0.0657 - acc: 1.0000
448/2960 [===>..........................] - ETA: 0s - loss: 0.0772 - acc: 0.9911
864/2960 [=======>......................] - ETA: 0s - loss: 0.0759 - acc: 0.9896
1312/2960 [============>.................] - ETA: 0s - loss: 0.0771 - acc: 0.9909
1824/2960 [=================>............] - ETA: 0s - loss: 0.0713 - acc: 0.9918
2304/2960 [======================>.......] - ETA: 0s - loss: 0.0689 - acc: 0.9913
2656/2960 [=========================>....] - ETA: 0s - loss: 0.0676 - acc: 0.9921
2960/2960 [==============================] - 0s 120us/step - loss: 0.0671 - acc: 0.9922 - val_loss: 0.0576 - val_acc: 0.9848
Epoch 16/20

32/2960 [..............................] - ETA: 0s - loss: 0.0534 - acc: 1.0000
512/2960 [====>.........................] - ETA: 0s - loss: 0.0768 - acc: 0.9863
1088/2960 [==========>...................] - ETA: 0s - loss: 0.0755 - acc: 0.9881
1536/2960 [==============>...............] - ETA: 0s - loss: 0.0706 - acc: 0.9902
1984/2960 [===================>..........] - ETA: 0s - loss: 0.0725 - acc: 0.9889
2400/2960 [=======================>......] - ETA: 0s - loss: 0.0713 - acc: 0.9892
2848/2960 [===========================>..] - ETA: 0s - loss: 0.0695 - acc: 0.9891
2960/2960 [==============================] - 0s 115us/step - loss: 0.0688 - acc: 0.9892 - val_loss: 0.0506 - val_acc: 0.9909
Epoch 17/20

32/2960 [..............................] - ETA: 0s - loss: 0.0731 - acc: 1.0000
256/2960 [=>............................] - ETA: 0s - loss: 0.0653 - acc: 0.9961
576/2960 [====>.........................] - ETA: 0s - loss: 0.0565 - acc: 0.9948
896/2960 [========>.....................] - ETA: 0s - loss: 0.0589 - acc: 0.9911
1312/2960 [============>.................] - ETA: 0s - loss: 0.0711 - acc: 0.9855
1856/2960 [=================>............] - ETA: 0s - loss: 0.0668 - acc: 0.9860
2240/2960 [=====================>........] - ETA: 0s - loss: 0.0662 - acc: 0.9862
2656/2960 [=========================>....] - ETA: 0s - loss: 0.0658 - acc: 0.9864
2960/2960 [==============================] - 0s 140us/step - loss: 0.0646 - acc: 0.9875 - val_loss: 0.0512 - val_acc: 0.9878
Epoch 18/20

32/2960 [..............................] - ETA: 0s - loss: 0.0258 - acc: 1.0000
672/2960 [=====>........................] - ETA: 0s - loss: 0.0524 - acc: 0.9911
1216/2960 [===========>..................] - ETA: 0s - loss: 0.0554 - acc: 0.9910
1664/2960 [===============>..............] - ETA: 0s - loss: 0.0576 - acc: 0.9904
2240/2960 [=====================>........] - ETA: 0s - loss: 0.0590 - acc: 0.9888
2912/2960 [============================>.] - ETA: 0s - loss: 0.0635 - acc: 0.9880
2960/2960 [==============================] - 0s 94us/step - loss: 0.0647 - acc: 0.9878 - val_loss: 0.0447 - val_acc: 0.9939
Epoch 19/20

32/2960 [..............................] - ETA: 0s - loss: 0.0354 - acc: 1.0000
480/2960 [===>..........................] - ETA: 0s - loss: 0.0509 - acc: 0.9958
928/2960 [========>.....................] - ETA: 0s - loss: 0.0623 - acc: 0.9881
1472/2960 [=============>................] - ETA: 0s - loss: 0.0614 - acc: 0.9885
1984/2960 [===================>..........] - ETA: 0s - loss: 0.0594 - acc: 0.9889
2336/2960 [======================>.......] - ETA: 0s - loss: 0.0592 - acc: 0.9893
2688/2960 [==========================>...] - ETA: 0s - loss: 0.0590 - acc: 0.9896
2912/2960 [============================>.] - ETA: 0s - loss: 0.0604 - acc: 0.9894
2960/2960 [==============================] - 0s 149us/step - loss: 0.0602 - acc: 0.9895 - val_loss: 0.0434 - val_acc: 0.9878
Epoch 20/20

32/2960 [..............................] - ETA: 0s - loss: 0.0393 - acc: 1.0000
416/2960 [===>..........................] - ETA: 0s - loss: 0.0702 - acc: 0.9832
768/2960 [======>.......................] - ETA: 0s - loss: 0.0632 - acc: 0.9870
1344/2960 [============>.................] - ETA: 0s - loss: 0.0577 - acc: 0.9888
1888/2960 [==================>...........] - ETA: 0s - loss: 0.0533 - acc: 0.9899
2464/2960 [=======================>......] - ETA: 0s - loss: 0.0515 - acc: 0.9911
2960/2960 [==============================] - 0s 118us/step - loss: 0.0513 - acc: 0.9916 - val_loss: 0.0416 - val_acc: 0.9878

Layer (type) Output Shape Param #

dense_1 (Dense) (None, 128) 4736

batch_normalization_1 (Batch (None, 128) 512

dense_2 (Dense) (None, 64) 8256

batch_normalization_2 (Batch (None, 64) 256

dense_3 (Dense) (None, 16) 1040

batch_normalization_3 (Batch (None, 16) 64

dense_4 (Dense) (None, 4) 68

Total params: 14,932
Trainable params: 14,516
Non-trainable params: 416

Traceback (most recent call last):
File "H:/Online-Realtime-Action-Recognition-based-on-OpenPose-master/Online-Realtime-Action-Recognition-based-on-OpenPose-master/Action/training/train.py", line 143, in
his.loss_plot('epoch')
File "H:/Online-Realtime-Action-Recognition-based-on-OpenPose-master/Online-Realtime-Action-Recognition-based-on-OpenPose-master/Action/training/train.py", line 55, in loss_plot
iters = range(len(self.losses[loss_type]))
AttributeError: 'LossHistory' object has no attribute 'losses'

Process finished with exit code 1

When I run main.py,the following error occurred. How did I solve it?
OSError: Unable to open file (unable to open file: name = 'Action/framewise_recognition.h5', errno = 2, error message = 'No such file or directory', flags = 0, o_flags = 0)

I have trained the model using only one action.Training for one action is successfully done.While testing , the webcam is running for about 2 seconds and it is getting terminated.And I got this error.

Awesome job, I was wondering if you could tell me the differences between framewise_recognition.h5 and framewise_recognition_under_scene.h5. I tried both and the second one seems more accurate. it is right?. Also I want to add more poses to your dataset, how can I these other poses?. It is possible to simulate amputations or bumps?. I was thinking to do this on Unity but no clue.

What is your platform configuration？such as GPU and CPU

Hello, Can you specify what type of License is provided?

THank you!

The action recognizer has a good performance on labeled clip but poor on webcam.

Hi @LZQthePlane does this include the feature to identity person name/ID?