Predicting Depth, Surface Normals and Semantic Labels with a Common Multi-scale Convolutional Architecture.

David Eigen Rob Fergus

Published in: ICCV (2015)

Keyphrases

surface normals
multiscale
semantic labels
systematic errors
photometric stereo
shape from shading
light source
surface reconstruction
shape recovery
scale space
image representation
statistical model
feature selection
multi view
viewpoint
visual content
high level
three dimensional