Paper Info
Title
Refine And Distill: Exploiting Cycle-Inconsistency And Knowledge Distillation For Unsupervised Monocular Depth Estimation
Abstract
Nowadays, the majority of state of the art monocular depth estimation techniques are based on supervised deep learning models. However, collecting RGB images with associated depth maps is a very time consuming procedure. Therefore, recent works have proposed deep architectures for addressing the monocular depth prediction task as a reconstruction problem, thus avoiding the need of collecting ground-truth depth. Following these works, we propose a novel self-supervised deep model for estimating depth maps. Our framework exploits two main strategies: refinement via cycle-inconsistency and distillation. Specifically, first a student network is trained to predict a disparity map such as to recover from a frame in a camera view the associated image in the opposite view. Then, a backward cycle network is applied to the generated image to re-synthesize back the input image, estimating the opposite disparity. A third network exploits the inconsistency between the original and the reconstructed input frame in order to output a refined depth map. Finally, knowledge distillation is exploited, such as to transfer information from the refinement network to the student. Our extensive experimental evaluation demonstrate the effectiveness of the proposed framework which outperforms state of the art unsupervised methods on the KITTI benchmark.
Year
DOI
Venue
2019
10.1109/CVPR.2019.01000
2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019)
Field
DocType
Volume
Reconstruction problem,Pattern recognition,Computer science,Exploit,Associated image,Distillation,RGB color model,Artificial intelligence,Deep learning,Depth map,Monocular
Journal
abs/1903.04202
ISSN
Citations 
PageRank 
1063-6919
7
0.40
References 
Authors
19
4
Name
Order
Citations
PageRank
Andrea Pilzer1292.74
Stéphane Lathuilière2335.98
Nicu Sebe37013403.03
Elisa Ricci 00024139373.75