Hi what does ``screen_w`` and ``screen_h`` mean in SemanticMapping?

[Yonggie]

In model.py, this following code computes camera instrinsic parameters.

And based on the get camera matrix code, it seems that the screen_w being the width of cmos and so the screen_h? screen_h is 120 and screen_w is 160, which also exactly the same as downscaled depth shape. In this way, is the screen_w has double meaning as both the shape of depth and cmos shape?

        self.camera_matrix = du.get_camera_matrix(
            self.screen_w, self.screen_h, self.fov)