video cropping

[kingkw1]

I've encountered issues with using the detection_box argument in the Video.analyze() function. With a little digging, I've found that self._crop(frame, detection_box) results in the following error:

{TypeError}_crop() missing 1 required positional argument: 'detection_box'

This problem stems from the combination of a @staticmethod decorator and "self" as an input argument to the _crop() function, and can be resolved by removing the "self" input argument from the _crop function definition.

    @staticmethod
    def _crop(self, frame, detection_box):

For the time being I've gotten around this with a monkey patch:

from fer import Video
Video._crop = lambda self, frame, _detection_box: frame[
                                                  _detection_box.get("y_min"): _detection_box.get("y_max"),
                                                  _detection_box.get("x_min"): _detection_box.get("x_max")]

[JustinShenk]

Thank you for opening this! This has been fixed in v22.0.1.