Error in analyzing videos: ValueError: could not broadcast input array from shape (8,24) into shape (4,24)

Hello, I am working on a MacOS to track 8 body parts on a single animal. I have analyzed and generated labeled videos for a large batch (~200) videos, however for about 10 of the videos the video analysis step generates the following error code:

ValueError: could not broadcast input array from shape (8,24) into shape (4,24)

I have tried converting the video file to grayscale and also adding the argument TFGPUinference=True – the error code continues to appear. Are there any other ways to resolve this?

Thanks!

you may be out of memory for your file size, try to downsample the videos (see dlc.DownSample for help!)

@vwong626, could you please post the full error traceback? Knowing where exactly in the code the error happened helps us tremendously :slight_smile: