FrameExporter

plugin FrameExporter(Predictor)

Plugin Type: Predictor

Exports probability maps to a binary format that can be passed back into DeepLabCut again to perform frame predictions later. This allows for a video to be run through the neural network (expensive) on a headless server or supercomputer, and then run through a predictor with gui feedback on a laptop or somewhere else.

Settings

setting sparsify: bool = True

Specify whether to optimize and store the data in a sparse format when dumping frames.

setting threshold: RangedFloat[min=0.0, max=1.0] = 1e-07

The threshold used if sparsify is true. Any values which land below this threshold probability won’t be included in the frame.

setting compression_level: RangedInteger[min=0.0, max=9.0] = 6

Determines the z-lib compression level. Higher compression level means it takes longer to compress the data, while 0 is no compression. Note this only applies if the dlfs format is being used, the hdf5 format ignores this value.

setting filename_suffix: str = 'DATA'

A string, The suffix to place onto the end of the file name.

setting bodyparts_to_keep: Literal[None] | List[str] = None

A list of body parts to store. None means keep all the body parts.

setting include_video: bool = True

If true, the video is embedded in the file, making this file fully independent. Defaults to True.