I am trying to decipher the ParkNet DNN introduced by Nvidia. With no research paper released, this is proving quite a hassle. Right at the end of the blog post, it is mentioned that, using the roll pitch and yaw values of the camera, the results from ParkNet are converted into 3D coordinates.
I have provided a link to the blog post below as well as the quote to which I am referring. Any help would be appreciated.
"ParkNet outputs parking space detections and entry line classifications in 2D image space. So a change from 2D to 3D coordinates is needed to use ParkNet outputs in autonomous parking planning and control software."
"By using camera self-calibration results (that is, estimates for the camera pitch/yaw/roll values, that represent up/down, left/right, and clockwise/anticlockwise positioning), ParkNet results can be converted into 3D coordinates. This enables 3D position estimation that’s particularly accurate for short-distance self-parking maneuvers."