Mick West
Mick West
social1h ago𝕏
Garry is wrong here.
Garry is wrong here. Firstly, the root problem is that Nandal and Loeb used poorly conditioned data. Their tight cluster of points is super sensitive to small pixel inaccuracies. Thousands of meters from a few pixels. The result they got for height was essentially random. Secondly, the result of that is that IT DOES NOT FIT THE DATA - specifically the video, which is the only real data we have. The angle is wrong, the horizon is off the screen, the terrain does not line up. Put the camera in...
