US Patent Application 17827019. WARPING DEPTH FEATURES FOR DEPTH ESTIMATION simplified abstract
Contents
WARPING DEPTH FEATURES FOR DEPTH ESTIMATION
Organization Name
TOYOTA JIDOSHA KABUSHIKI KAISHA
Inventor(s)
Vitor Guizilini of Santa Clara CA (US)
Rares A. Ambrus of San Francisco CA (US)
Sergey Zakharov of San Francisco CA (US)
WARPING DEPTH FEATURES FOR DEPTH ESTIMATION - A simplified explanation of the abstract
This abstract first appeared for US patent application 17827019 titled 'WARPING DEPTH FEATURES FOR DEPTH ESTIMATION
Simplified Explanation
- The patent application describes an improved approach to training a depth model for monocular depth estimation. - The method involves encoding a source image into depth features using an encoder of the depth model. - The depth features are then warped into warped features of a target frame of a target image associated with the source image. - The warped features are decoded into a depth map using a decoder of the depth model. - The depth model is trained based on a loss derived from the depth map.
Original Abstract Submitted
System, methods, and other embodiments described herein relate to an improved approach to training a depth model for monocular depth estimation by warping depth features prior to decoding. In one embodiment, a method includes encoding, using an encoder of a depth model, a source image into depth features of a scene depicted by the source image. The method includes warping the depth features into warped features of a target frame of a target image associated with the source image. The method includes decoding, using a decoder of the depth model, the warped features into a depth map. The method includes training the depth model according to a loss derived from the depth map.
