labelme data_annotated --labels labels.txt --nodata --keep-prev --config '{shift_auto_shape_color: -2}'
Fig 1. Video annotation example. A frame (left), The next frame (right).
Fig 2. Visualization of video semantic segmentation.
pip install video-cli
video-toimg your_video.mp4 # this creates your_video/ directory
ls your_video/
labelme your_video/