If it worked, continue with the next section. Otherwise:
- If these failed with an out of memory error, check and follow [Improving Memory and Speed but Decreasing Accuracy](#improving-memory-and-speed-but-decreasing-accuracy).
- If you are using Mac, make sure to check and follow [Mac OSX Additional Step](#mac-osx-additional-step).
- Otherwise, check [FAQ](#faq).
- If these failed with an out of memory error, check and follow the section [Improving Memory and Speed but Decreasing Accuracy](#improving-memory-and-speed-but-decreasing-accuracy).
- If you are using Mac, make sure to check and follow the section [Mac OSX Additional Step](#mac-osx-additional-step).
@@ -68,7 +68,7 @@ The output of the JSON files consist of a set of keypoints, whose ordering is re
## JSON Output Format
There are 2 alternatives to save the OpenPose output. But both of them follow the keypoint ordering described in the [Keypoint Ordering in C++/Python](#body-keypoints-in-c-python) section (which you should read next).
There are 2 alternatives to save the OpenPose output. But both of them follow the keypoint ordering described in the the section [Keypoint Ordering in C++/Python](#body-keypoints-in-c-python) section (which you should read next).
1. The `--write_json` flag saves the people pose data onto JSON files. Each file represents a frame, it has a `people` array of objects, where each object has:
1.`pose_keypoints_2d`: Body part locations (`x`, `y`) and detection confidence (`c`) formatted as `x0,y0,c0,x1,y1,c1,...`. The coordinates `x` and `y` can be normalized to the range [0,1], [-1,1], [0, source size], [0, output size], etc. (see the flag `--keypoint_scale` for more information), while the confidence score (`c`) in the range [0,1].
For quick prototyping, you can simply duplicate and rename any of the existing sample files in `build/examples/tutorial_api_python` within that same folder and start building in there. These files are copied from [existing example files](../examples/tutorial_api_python/) on compiling time. 2 alternatives:
- You can either duplicate and create your files in [examples/tutorial_api_python/](../examples/tutorial_api_python/), but you will have to recompile OpenPose every time you make changes to your Python files so they are copied over the `build/` folder.
For quick prototyping, you can simply duplicate and rename any of the existing sample files in `build/examples/tutorial_api_python` within that same folder and start building in there. These files are copied from [existing example files](https://github.com/CMU-Perceptual-Computing-Lab/openpose/tree/master/examples/tutorial_api_python) on compiling time. 2 alternatives:
- You can either duplicate and create your files in [examples/tutorial_api_python/](https://github.com/CMU-Perceptual-Computing-Lab/openpose/tree/master/examples/tutorial_api_python), but you will have to recompile OpenPose every time you make changes to your Python files so they are copied over the `build/` folder.
- Or you can directly edit them in `build/examples/tutorial_api_python`. This does not require rebuilding, but cleaning OpenPose will remove the whole `build/` folder, so make sure to back your files up!
@@ -30,7 +30,7 @@ OpenPose displays the FPS in the basic GUI. However, more complex speed metrics
- Time measurement for 1 graphic card: The FPS will be the slowest time displayed in your terminal command line (as OpenPose is multi-threaded). Times are in milliseconds, so `FPS = 1000/millisecond_measurement`.
- Time measurement for >1 graphic cards: Assuming `n` graphic cards, you will have to wait up to `n` x `F` frames to visualize each graphic card speed (as the frames are splitted among them). In addition, the FPS would be: `FPS = minFPS(speed_per_GPU/n, worst_time_measurement_other_than_GPUs)`. For < 4 GPUs, this is usually `FPS = speed_per_GPU/n`.
Make sure that `wPoseExtractor` time is the slowest timing. Otherwise the input producer (video/webcam codecs issues with OpenCV, images too big, etc.) or the GUI display (use OpenGL support as detailed in [Speed Up Preserving Accuracy](#speed-up-preserving-accuracy)) might not be optimized.
Make sure that `wPoseExtractor` time is the slowest timing. Otherwise the input producer (video/webcam codecs issues with OpenCV, images too big, etc.) or the GUI display (use OpenGL support as detailed in the next section (`Speed Up Preserving Accuracy`) might not be optimized.
Here we expose all projects created with OpenPose by the community and that were shared with us. Do you want to share yours? Simply create a pull request and add to this file your demo and a description of it!
1.
2.[Add here your demo name and link](#here_some_full_link): Add here the description of your project.
2.[**Hand gesture classification application - OpenHand**](https://github.com/ArthurFDLR/OpenHand-App): Third-party application that eases hand keypoints datasets creation and real-time hand gesture classification. You can deploy your own Neural Network classification model on top of OpenPose and play with it in real-time through a GUI!
3.[Add here your demo name and link](#here_some_full_link): Add here the description of your project.
Disclaimer: We do not support any of these projects, we are simply exposing them. GitHub issues or questions about those will result in strict user bans and the posts being deleted.
@@ -82,7 +82,7 @@ There is an exception, for CPU version, the COCO and MPI models seems to be fast
## Help Flag
We recommend [All Flags](#all-flags) shows them all in this document, which sorts all the flags by category.
We recommend following the next section (`All Flags`), which shows all the flags in this document and sorts them by category.
However, you could add the flag `--help` at any point to see all the available OpenPose flags. Check only the flags for `examples/openpose/openpose.cpp` itself (i.e., the ones in the section `Flags from examples/openpose/openpose.cpp:`).
4. Then, you can run OpenPose from the PowerShell command-line by following [doc/01_demo.md](../01_demo.md).
Note: If you are using the GPU-accelerated version and are seeing `Cuda check failed (3 vs. 0): initialization error` when running OpenPose, you can fix it by doing one of these:
- Upgrade your Nvidia drivers. If the error persists, make sure your machine does not contain any CUDA version (or if so, that it's the same than the OpenPose portable demo files). Otherwise, uninstall that CUDA version. If you need to keep that CUDA version installed, [compile OpenPose from Source](#compiling-and-running-openpose-from-source) for that CUDA version instead.
- Upgrade your Nvidia drivers. If the error persists, make sure your machine does not contain any CUDA version (or if so, that it's the same than the OpenPose portable demo files). Otherwise, uninstall that CUDA version. If you need to keep that CUDA version installed, follow [Compiling and Running OpenPose from Source](#compiling-and-running-openpose-from-source) for that particular CUDA version instead.
- Download an older OpenPose version (v1.6.0 does not show this error).
...
...
@@ -221,7 +221,7 @@ Check OpenPose was properly installed by running any demo example: [doc/01_demo.
### Custom User Code
You can quickly add your custom code so that quick prototypes can be easily tested without having to create a whole new project just for it. See [examples/user_code/README.md](../../examples/user_code/README.md) for more details.
You can quickly add your custom code so that quick prototypes can be easily tested without having to create a whole new project just for it. See [examples/user_code/README.md](https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/examples/user_code/README.md) for more details.
@@ -44,12 +44,11 @@ Recommended installation method, it is simpler and offers more customization set
### Prerequisites (Script Compilation or Manual Compilation)
CUDA, cuDNN, OpenCV and Atlas must be already installed on your machine:
1. [CUDA](https://developer.nvidia.com/cuda-80-ga2-download-archive) must be installed. You should reboot your machine after installing CUDA.
2. [cuDNN](https://developer.nvidia.com/cudnn): Once you have downloaded it, just unzip it and copy (merge) the contents on the CUDA folder, e.g., `/usr/local/cuda-8.0/`. Note: We found OpenPose working ~10% faster with cuDNN 5.1 compared to cuDNN 6. Otherwise, check [Compiling without cuDNN](#compiling-without-cudnn).
3. OpenCV can be installed with `apt-get install libopencv-dev`. If you have compiled OpenCV 3 by your own, follow [Manual Compilation](#manual-compilation). After both Makefile.config files have been generated, edit them and uncomment the line `# OPENCV_VERSION := 3`. You might alternatively modify all `Makefile.config.UbuntuXX` files and then run the scripts in step 2.
4. In addition, OpenCV 3 does not incorporate the `opencv_contrib` module by default. Assuming you have OpenCV 3 compiled with the contrib module and you want to use it, append `opencv_contrib` at the end of the line `LIBRARIES += opencv_core opencv_highgui opencv_imgproc` in the `Makefile` file.
5. Atlas can be installed with `sudo apt-get install libatlas-base-dev`. Instead of Atlas, you can use OpenBLAS or Intel MKL by modifying the line `BLAS := atlas` in the same way as previosuly mentioned for the OpenCV version selection.
1.[CUDA](https://developer.nvidia.com/cuda-80-ga2-download-archive) must be installed. You should reboot your machine after installing CUDA.
2.[cuDNN](https://developer.nvidia.com/cudnn): Once you have downloaded it, just unzip it and copy (merge) the contents on the CUDA folder, e.g., `/usr/local/cuda-8.0/`. Note: We found OpenPose working ~10% faster with cuDNN 5.1 compared to cuDNN 6. Otherwise, check the section [Compiling without cuDNN](#compiling-without-cudnn).
3. OpenCV can be installed with `apt-get install libopencv-dev`. If you have compiled OpenCV 3 by your own, follow [Manual Compilation](#manual-compilation). After both Makefile.config files have been generated, edit them and uncomment the line `# OPENCV_VERSION := 3`. You might alternatively modify all `Makefile.config.UbuntuXX` files and then run the scripts in step 2.
4. In addition, OpenCV 3 does not incorporate the `opencv_contrib` module by default. Assuming you have OpenCV 3 compiled with the contrib module and you want to use it, append `opencv_contrib` at the end of the line `LIBRARIES += opencv_core opencv_highgui opencv_imgproc` in the `Makefile` file.
5. Atlas can be installed with `sudo apt-get install libatlas-base-dev`. Instead of Atlas, you can use OpenBLAS or Intel MKL by modifying the line `BLAS := atlas` in the same way as previosuly mentioned for the OpenCV version selection.
3. Smart multi-threading: Some classes are much more faster than others (e.g., pose estimation takes ~100 ms while extracting frames from a video only ~10 ms). In addition, any machine has a limited number of threads. Therefore, the library allows the user to merge the faster threads in order to potentially speed up the code. Check the [real-time pose demo](../../../examples/openpose/openpose.cpp) too see a more complete example.
3. Smart multi-threading: Some classes are much more faster than others (e.g., pose estimation takes ~100 ms while extracting frames from a video only ~10 ms). In addition, any machine has a limited number of threads. Therefore, the library allows the user to merge the faster threads in order to potentially speed up the code. Check the [command line demo](https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/examples/openpose/openpose.cpp) too see a more complete example.
