You need dependencies below.

• python3
• tensorflow 1.4.1+
• opencv3, protobuf, python3-tk
• slidingwindow
  • https://github.com/adamrehn/slidingwindow
  • I copied from the above git repo to modify few things.

Personally, I used the next dependencies:

Dedicate use of GPU:

• Cuda v.10
• cuDNN v7.4 Other dependencies:
• python3
• tensorflow 1.4.1

Operative System:

• Windows 10

Clone the repo and install 3rd-party libraries.

$ git clone https://github.com/paulnajera/skeleton_humanactivity.git
$ cd skeleton_humanactivity
$ pip3 install -r requirements.txt

If there is a problem installing the requiremets (related to MVS and pycocotools) I followed this steps and it worked: https://www.kaggle.com/c/tgs-salt-identification-challenge/discussion/62381

Build c++ library for post processing. See : https://github.com/ildoonet/tf-pose-estimation/tree/master/tf_pose/pafprocess (wig have to be installed in order to run this command)

$ cd tf_pose/pafprocess
$ swig -python -c++ pafprocess.i && python3 setup.py build_ext --inplace

See experiments.md

Before running, you should download graph files. You can deploy this graph on your mobile or other platforms.

• cmu (trained in 656x368)
• mobilenet_thin (trained in 432x368)
• mobilenet_v2_large (trained in 432x368)
• mobilenet_v2_small (trained in 432x368)

CMU's model graphs are too large for git, so I uploaded them on an external cloud. You should download them if you want to use cmu's original model. Download scripts are provided in the model folder.

$ cd models/graph/cmu
$ bash download.sh

In order to test the next steps, the model have to be downloaded.

You can test the inference feature with a single image.

$ python run.py --model=mobilenet_thin --resize=432x368 --image=./images/p1.jpg

The image flag MUST be relative to the src folder with no "~", i.e:

--image ../../Desktop

Then you will see the screen as below with pafmap, heatmap, result and etc.

$ python run_webcam.py

Including other options:

$ python run_webcam.py --model=mobilenet_thin --resize=432x368 --camera=0

Apply TensoRT

$ python run_webcam.py --model=mobilenet_thin --resize=432x368 --camera=0 --tensorrt=True

$ python run_video.py --video=videos/test3.mp4

In both cases, the result is recorded and save in the main folder where the .py are)

$ python webcam_3D.pz

$ python video_3D.py

Video source shoul be specified in line 50:

camera = './skeleton_humanactivity/videos/test1-2.mp4'

This pose estimator provides simple python classes that you can use in your applications.

See run.py or run_webcam.py as references.

e = TfPoseEstimator(get_graph_path(args.model), target_size=(w, h))
humans = e.inference(image)
image = TfPoseEstimator.draw_humans(image, humans, imgcopy=False)

See : etcs/ros.md

See : etcs/training.md

See : etcs/reference.md