There are lots of crossing lines, if the AI took those geometrical shapes into account it should have placed lines in front of his eyes and mouth. That would ha

There are lots of crossing lines, if the AI took those geometrical shapes into account it should have placed lines in front of his eyes and mouth. That would have been more accourate, incidentally avoiding the match with Goslin. Is that possible to implement in Gigapixel? Asking as a CV enthusiast that tailored a volume recognition solution in OpenCV years ago.