search

icatcherplus / icatcher_plus

iCatcher+: Robust and automated annotation of infant gaze from videos collected in laboratory, field, and online studies
GNU General Public License v3.0
18 stars 19 forks source link

Mislabeling of classes in visualize.py output #41

Closed samwinebrake closed 1 year ago

samwinebrake commented 1 year ago

Not a big deal, just noticed that in the text output of reproduce/visualize.py, the invalid frame breakdown is based off of classes formed by:

invalid_no_face += np.count_nonzero(invalids == -1)
invalid_no_infant_face += np.count_nonzero(invalids == -2)

This should just be changed if visualize is to be used in future assessments. (Just flipping the numbers, no_face should align w/ -2s, and no_infant_face should align with -1s since classes = {'noface': -2, 'nobabyface': -1, 'away': 0, 'left': 1, 'right': 2})

invalid_no_face += np.count_nonzero(invalids == -2)
invalid_no_infant_face += np.count_nonzero(invalids == -1)