Skip to content

Latest commit

 

History

History
86 lines (59 loc) · 1.32 KB

README.md

File metadata and controls

86 lines (59 loc) · 1.32 KB
Raw

Local setup

Note
  • For non-GPU users, please install CPU version of PyTorch first
pip install -i https://download.pytorch.org/whl/cpu torch torchvision

pip install -r requirements.txt
streamlit run app.py

or run in docker

docker-compose up

Features

  • Run locally on web UI

    • Model

      • Object detection
      • Object segmentation
      • Pose estimation
      • Image classification

    • On

      • Image
      • Video
      • Webcam

    • With ability to

      • Turn tracking on/off
      • Adjust confidence threshold
      • Filter by class
      • Object motion path
      • Object color classification
      • Trim video

  • Draw visual elements interactively

    • Line count (in/out)
    • Polygon zone count

  • Customize visual elements

    • Toggle on/off

      • Box
      • Label
      • Mask
      • Area
      • Trail
      • Count
      • FPS

    • Adjust

      • Text size
      • Text color
      • Text padding
      • Text offset
      • Line thickness
      • Mask opacity
      • Trail length

  • PRODUCTION READY

    • Save drawn visual elements & settings in JSON
    • Run inference with OpenCV standalone from saved JSON
Note

Camera (/dev/video0) & native run (cv2.imshow()) is not configured to run in docker (you can try to mount your own device)