Method and device for learning a depth prediction model of a set of pixels of an image associated with a stereoscopic vision system embedded in a vehicle.

Abstract

A method or device for learning a depth prediction model implemented by a neural network associated with a vision system. The method comprises the generation (32) of a set of feature map pairs from images acquired by the vision system. Each feature map pair includes first and second feature maps associated with each acquired image, with each pair having a different definition. Another feature map pair is associated (35) with each feature pair and includes two additional feature maps generated from depths and directions (33, 34) predicted for the pixels of the first and second feature maps. The depth prediction model is learned by minimizing (37) an error determined by comparing the feature maps for different resolutions.Figure for the abbreviation: Figure 3.

Images