Closed 408550969 closed 4 months ago
If I perform landmark detection on each frame of the image and then expand the corresponding bounding matrix by a certain factor before sending it for training, can this effectively overcome moving scenes?
If I perform landmark detection on each frame of the image and then expand the corresponding bounding matrix by a certain factor before sending it for training, can this effectively overcome moving scenes?