Cited By
View all- Li YWei XFan H(2023)Attention Mechanism Used in Monocular Depth Estimation: An OverviewApplied Sciences10.3390/app1317994013:17(9940)Online publication date: 2-Sep-2023
Monocular depth estimation using synthetic data with domain-separated feature alignment
Authors: 
Guan Wang, Yundong LiAuthors Info & Claims
Guan Wang, Yundong LiAuthors Info & ClaimsPages 100 - 105
Abstract
Depth estimation is a research hotspot in today's computer vision tasks, and its depth information is of great significance for many applications such as autonomous driving, 3D reconstruction, and object tracking. Although supervised depth estimation based on deep learning has high prediction accuracy, it is difficult to obtain a large amount of labeled data. Therefore, self-supervised monocular depth estimation methods that do not require labeled data have become the mainstream of research. Self-supervised learning utilizes the constraints of objects on spatial geometry, reducing the need for labeled data, but the problems of dynamic objects, shooting angles, and visibility weaken the effect of self-supervision. Using the accurate depth information of synthetic datasets to assist the self-supervised training of real datasets can improve the accuracy of self-supervised depth estimation, but most methods do not consider the distribution difference between synthetic data and real data, which affects the estimation effect. Aiming at this problem, this paper proposes a domain-separated Monocular Depth Estimation (DsMDE) algorithm based on domain separation network, which uses orthogonal loss to separate the public and private features of each domain, and then uses the maximum mean difference to The common features are aligned to reduce the difference between the synthetic domain and the real domain. The experimental results show that the DsMDE method proposed in this paper improves the depth estimation effect, and the depth estimation accuracy is better than that of the mainstream algorithms.
References
[1]
Eigen D, Puhrsch C, Fergus R. Depth map prediction from a single image using a multi-scale deep network[J]. In Advances in neural information processing systems,2014:2366-2374. https://arxiv.org/abs/1406.2283
[2]
Godard, C.; Mac Aodha, O.; Firman, M.; Brostow, G.J. Digging into self-supervised monocular depth estimation. In Proceedings of the 2019 IEEE International Conference on Computer Vision, Seoul, Korea,27–28 October 2019; pp. 3828–3838. https://arxiv.org/abs/1806.01260
[3]
Laina I, Rupprecht C, Belagiannis V, Deeper Depth Prediction with Fully Convolutional Residual Networks[J]. Fourth International Conference on 3d Vision, 2016. https://arxiv.org/abs/1606.00373v1
[4]
Fu H, Gong M, Wang C, Deep Ordinal Regression Network for Monocular Depth Estimation[C]// 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, 2018. https://arxiv.org/abs/1806.02446
[5]
Wofk D, Ma F, Yang T J, FastDepth: Fast Monocular Depth Estimation on Embedded Systems[J]. IEEE, 2019. https://arxiv.org/abs/1903.03273
[6]
Zhou T, Brown M, Snavely N, Unsupervised Learning of Depth and Ego-Motion from Video[J]. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017. http://arxiv.org/abs/1704.07813
[7]
Yin Z, Shi J . GeoNet: Unsupervised Learning of Dense Depth, Optical Flow and Camera Pose[C]// 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2018. http://arxiv.org/abs/1803.02276
[8]
N. Mayer, E. Ilg, P. Hausser, P. Fischer, D. Cremers, A. Dosovitskiy, T. Brox, A large dataset to train convolutional networks for disparity, opticalflow, and sceneflow estimation, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016, pp. 4040–4048. https://arxiv.org/abs/1512.02134v1
[9]
Zheng C, Cham T J, Cai J . T2Net: Synthetic-to-Realistic Translation for Solving Single-Image Depth Estimation Tasks[J]. Springer, Cham, 2018. pp. 767–783. http://arxiv.org/abs/1808.01454
[10]
B. Cheng, I. S. Saggu, R. Shah, G. Bansal, and D. Bharadia,“S3Net:Semantic-aware self-supervised depth estimation with monocular videos and synthetic data,” in Proc. Eur. Conf. 20. http://arxiv.org/abs/2007.14511
[11]
Pnvr K, Zhou H, Jacobs D . SharinGAN: Combining Synthetic and Real Data for Unsupervised Geometry Estimation[C]// 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2020. http://arxiv.org/abs/2006.04026
[12]
Gurram A, Tuna A F, F Shen, Monocular Depth Estimation through Virtual-world Supervision and Real-world SfM Self-Supervision[J]. 2021. http://arxiv.org/abs/2103.12209v1
[13]
Liu L, Song X, Wang M, Self-supervised Monocular Depth Estimation for All Day Images using Domain Separation[J]. 2021. http://arxiv.org/abs/2108.07628
[14]
Godard C, Aodha O M, Brostow G J . Unsupervised Monocular Depth Estimation with Left-Right Consistency[C]// Computer Vision & Pattern Recognition. IEEE, 2017. http://arxiv.org/abs/1609.03677
[15]
Bousmalis K, Trigeorgis G, Silberman N, Domain Separation Networks[C]// 2016. http://arxiv.org/abs/1608.06019
Index Terms
- Monocular depth estimation using synthetic data with domain-separated feature alignment
Recommendations
Transferring knowledge from monocular completion for self-supervised monocular depth estimation
AbstractMonocular depth estimation is a very challenging task in computer vision, with the goal to predict per-pixel depth from a single RGB image. Supervised learning methods require large amounts of depth measurement data, which are time-consuming and ...
Net: Semantic-Aware Self-supervised Depth Estimation with Monocular Videos and Synthetic Data
Computer Vision – ECCV 2020AbstractSolving depth estimation with monocular cameras enables the possibility of widespread use of cameras as low-cost depth estimation sensors in applications such as autonomous driving and robotics. However, learning such a scalable depth estimation ...
Learning Monocular Depth by Distilling Cross-Domain Stereo Networks
Computer Vision – ECCV 2018AbstractMonocular depth estimation aims at estimating a pixelwise depth map for a single image, which has wide applications in scene understanding and autonomous driving. Existing supervised and unsupervised methods face great challenges. Supervised ...
Comments
Information & Contributors
Information
Published In
December 2022
341 pages
ISBN:9781450397773
DOI:10.1145/3577530
Copyright © 2022 ACM.
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 30 March 2023
Check for updates
Author Tags
Qualifiers
- Research-article
- Research
- Refereed limited
Funding Sources
- National Natural Science Foundation of China
Conference
CSAI 2022
CSAI 2022: 2022 6th International Conference on Computer Science and Artificial Intelligence
December 9 - 11, 2022
Beijing, China
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- View Citations1Total Citations
- 28Total Downloads
- Downloads (Last 12 months)7
- Downloads (Last 6 weeks)1
Reflects downloads up to 27 Feb 2025
Other Metrics
Citations
Cited By
View all- Li YWei XFan H(2023)Attention Mechanism Used in Monocular Depth Estimation: An OverviewApplied Sciences10.3390/app1317994013:17(9940)Online publication date: 2-Sep-2023
View Options
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign inFull Access
View options
View or Download as a PDF file.
PDFeReader
View online with eReader.
eReaderHTML Format
View this article in HTML Format.
HTML Format