Open cellistigs opened 2 years ago
I've added functions to segment the videos into n frames before running inference. I've also integrated this into the run_dgp_demo.py code, using two flags:
--split
is a boolean flag that tells the script to segment each video in the video_sets
variable. --splitlength
gives the length in frames of each video chunk.
I'm starting a pull request for the video segmentation we discussed. Nothing is implemented yet, but I've stubbed out the relevant functions in 'src/contrib/segment_videos.py'. Please feel free to add to this pr if you like.