Authors:
Yu Wang
and
Lizhuang Ma
Affiliation:
Shanghai Jiaotong University, China
Keyword(s):
Single-image Depth Prediction, CNN.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Computational Intelligence
;
Evolutionary Computing
;
Knowledge Discovery and Information Retrieval
;
Knowledge-Based Systems
;
Machine Learning
;
Soft Computing
;
Symbolic Systems
Abstract:
With the recent surge of deep neural networks, depth prediction
from a single image has seen substantial progress.
Deep regression networks are typically learned from large
data without much constraints about the scene structure, thus
often leading to uncertainties at discontinuous regions. In this
paper, we propose a structure-aware depth prediction method
based on two observations: depth is relatively smooth within
the same objects, and it is usually easier to model relative
depth than model the absolute depth from scratch. Our network
first predicts an initial depth map and takes an object
saliency map as input, which helps to teach the network to
learn depth refinement. Specifically, a stable anchor depth is
first estimated from the detected salient objects, and the learning
objective is to penalize the difference in relative depth versus
the estimated anchor.We show such saliency-guided relative
depth constraint unveils helpful scene structures, leading
to significant gains on t
he RGB-D saliency dataset NLPR and
depth prediction dataset NYU V2. Furthermore, our method
is appealing in that it is pluggable to any depth network and
is trained end-to-end with no overhead of time during testing.
Key words: Single-image Depth Prediction, CNN
(More)