CMU-Perceptual-Computing-Lab / openpose

OpenPose: Real-time multi-person keypoint detection library for body, face, hands, and foot estimation
https://cmu-perceptual-computing-lab.github.io/openpose
Other
30.33k stars 7.79k forks source link

Run successfully, but could not get face/hand frame #2269

Closed 1uan-yuan closed 6 months ago

1uan-yuan commented 6 months ago

Issue Summary

I could run OpenPose successfully, but I could not get the frame of hand and face correctly.

Executed Command (if any)

Note: add --logging_level 0 --disable_multi_thread to get higher debug information. bin\OpenPoseDemo.exe --video examples\media\video.avi --face --hand bin\OpenPoseDemo.exe --image_dir examples\media\ --face --hand

OpenPose Output (if any)

C:......\openpose_source>bin\OpenPoseDemo.exe --video examples\media\video.avi --face --hand Starting OpenPose demo... Configuring OpenPose... Starting thread(s)... Auto-detecting all available GPUs... Detected 2 GPU(s), using 2 of them starting at GPU 0. OpenPose demo successfully finished. Total time: 9.092402 seconds.

C:......\openpose_source>bin\OpenPoseDemo.exe --image_dir examples\media\ --face --hand Starting OpenPose demo... Configuring OpenPose... Starting thread(s)... Auto-detecting all available GPUs... Detected 2 GPU(s), using 2 of them starting at GPU 0. OpenPose demo successfully finished. Total time: 3.230856 seconds.

Errors (if any)

I could not get any of the data. Even though the program runs correctly, the frames of people are not detected. The following are the screenshots of what I could get:

截屏2023-12-19 下午6 33 02 截屏2023-12-19 下午6 33 50

I thought that there might be some bug exists in displaying. Therefore, I also tried to log out the data into a JSON file using command --write_json. The JSON files could be generated successfully, but all the data are null.

Type of Issue

Select the topic(s) on your post, delete the rest:

Your System Configuration

  1. Whole console output -C:......\openpose_source>bin\OpenPoseDemo.exe --video examples\media\video.avi --face --hand Starting OpenPose demo... Configuring OpenPose... Starting thread(s)... Auto-detecting all available GPUs... Detected 2 GPU(s), using 2 of them starting at GPU 0. OpenPose demo successfully finished. Total time: 9.092402 seconds.

    -C:......\openpose_source>bin\OpenPoseDemo.exe --image_dir examples\media\ --face --hand Starting OpenPose demo... Configuring OpenPose... Starting thread(s)... Auto-detecting all available GPUs... Detected 2 GPU(s), using 2 of them starting at GPU 0. OpenPose demo successfully finished. Total time: 3.230856 seconds.

  2. OpenPose version: The latest version from github.

  3. General configuration: I used CMake GUI to configure and used Visual Studio 16 2019 to build the sln file on Win10 system.

  4. Non-default settings:

    • 3-D Reconstruction module added? (by default: no): no
    • Any other custom CMake configuration with respect to the default version? (by default: no): no
  5. 3rd-party software:

    • Caffe version: I used the caffe contained in the post (https://github.com/CMU-Perceptual-Computing Lab/openpose/issues/1602).
    • CMake version cmake version 3.250-rc2
    • OpenCV version: pre-compiled apt-get install libopencv-dev (only Ubuntu); OpenPose default (only Windows); compiled from source? If so, 2.4.9, 2.4.12, 3.1, 3.2?; ...? opencv_450_v15_2020_11_18 provided in https://github.com/CMU-Perceptual-Computing Lab/openpose/issues/1602 .

I would really appreciate all kinds of help. Thank you! =)

1uan-yuan commented 6 months ago

Update:

I solved this problem. I think it might be due to the problem of wrong configuration or wrong build. I solved this problem by redo the whole procedure of installation. i.e., I downloaded OpenPose again and started with a new project.

utkarshshukla2912 commented 6 months ago

@1uan-yuan did you do the whole process without changing anything, what changes you made in the configuration

1uan-yuan commented 5 months ago

@1uan-yuan did you do the whole process without changing anything, what changes you made in the configuration

No, I did not change anything. I followed the exact same procedures.