UnDEMoN: Unsupervised Deep Network for Depth and Ego-Motion Estimation

V. Madhu Babu, Kaushik Das, Anima Majumdar, Swagat Kumar

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

This paper presents an unsupervised deep learning framework called UnDEMoN for estimating dense depth map and 6-DoF camera pose information directly from monocular images. The proposed network is trained using unlabeled monocular stereo image pairs and is shown to provide superior performance in depth and ego-motion estimation compared to the existing state-of-the-art. These improvements are achieved by introducing a new objective function that aims to minimize spatial as well as temporal reconstruction losses simultaneously. These losses are defined using bi-linear sampling kernel and penalized using the Charbonnier penalty function. The objective function, thus created, provides robustness to image gradient noises thereby improving the overall estimation accuracy without resorting to any coarse to fine strategies which are currently prevalent in the literature. Another novelty lies in the fact that we combine a disparity-based depth estimation network with a pose estimation network to obtain absolute scale-aware 6 DOF Camera pose and superior depth map. The effectiveness of the proposed approach is demonstrated through performance comparison with the existing supervised and unsupervised methods on the KITTI driving dataset.
Original languageEnglish
Title of host publicationIEEE International Conference on Intelligent Robots and Systems
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1082-1088
Number of pages7
ISBN (Print)9781538680940
DOIs
Publication statusPublished - 27 Dec 2018

Publication series

NameIEEE International Conference on Intelligent Robots and Systems

Fingerprint

Motion estimation
Cameras
Sampling

Cite this

Madhu Babu, V., Das, K., Majumdar, A., & Kumar, S. (2018). UnDEMoN: Unsupervised Deep Network for Depth and Ego-Motion Estimation. In IEEE International Conference on Intelligent Robots and Systems (pp. 1082-1088). (IEEE International Conference on Intelligent Robots and Systems). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IROS.2018.8593864
Madhu Babu, V. ; Das, Kaushik ; Majumdar, Anima ; Kumar, Swagat. / UnDEMoN: Unsupervised Deep Network for Depth and Ego-Motion Estimation. IEEE International Conference on Intelligent Robots and Systems. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 1082-1088 (IEEE International Conference on Intelligent Robots and Systems).
@inbook{fe9e6ef955a74998a50d9424a34aa459,
title = "UnDEMoN: Unsupervised Deep Network for Depth and Ego-Motion Estimation",
abstract = "This paper presents an unsupervised deep learning framework called UnDEMoN for estimating dense depth map and 6-DoF camera pose information directly from monocular images. The proposed network is trained using unlabeled monocular stereo image pairs and is shown to provide superior performance in depth and ego-motion estimation compared to the existing state-of-the-art. These improvements are achieved by introducing a new objective function that aims to minimize spatial as well as temporal reconstruction losses simultaneously. These losses are defined using bi-linear sampling kernel and penalized using the Charbonnier penalty function. The objective function, thus created, provides robustness to image gradient noises thereby improving the overall estimation accuracy without resorting to any coarse to fine strategies which are currently prevalent in the literature. Another novelty lies in the fact that we combine a disparity-based depth estimation network with a pose estimation network to obtain absolute scale-aware 6 DOF Camera pose and superior depth map. The effectiveness of the proposed approach is demonstrated through performance comparison with the existing supervised and unsupervised methods on the KITTI driving dataset.",
author = "{Madhu Babu}, V. and Kaushik Das and Anima Majumdar and Swagat Kumar",
year = "2018",
month = "12",
day = "27",
doi = "10.1109/IROS.2018.8593864",
language = "English",
isbn = "9781538680940",
series = "IEEE International Conference on Intelligent Robots and Systems",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "1082--1088",
booktitle = "IEEE International Conference on Intelligent Robots and Systems",
address = "United States",

}

Madhu Babu, V, Das, K, Majumdar, A & Kumar, S 2018, UnDEMoN: Unsupervised Deep Network for Depth and Ego-Motion Estimation. in IEEE International Conference on Intelligent Robots and Systems. IEEE International Conference on Intelligent Robots and Systems, Institute of Electrical and Electronics Engineers Inc., pp. 1082-1088. https://doi.org/10.1109/IROS.2018.8593864

UnDEMoN: Unsupervised Deep Network for Depth and Ego-Motion Estimation. / Madhu Babu, V.; Das, Kaushik; Majumdar, Anima; Kumar, Swagat.

IEEE International Conference on Intelligent Robots and Systems. Institute of Electrical and Electronics Engineers Inc., 2018. p. 1082-1088 (IEEE International Conference on Intelligent Robots and Systems).

Research output: Chapter in Book/Report/Conference proceedingChapter

TY - CHAP

T1 - UnDEMoN: Unsupervised Deep Network for Depth and Ego-Motion Estimation

AU - Madhu Babu, V.

AU - Das, Kaushik

AU - Majumdar, Anima

AU - Kumar, Swagat

PY - 2018/12/27

Y1 - 2018/12/27

N2 - This paper presents an unsupervised deep learning framework called UnDEMoN for estimating dense depth map and 6-DoF camera pose information directly from monocular images. The proposed network is trained using unlabeled monocular stereo image pairs and is shown to provide superior performance in depth and ego-motion estimation compared to the existing state-of-the-art. These improvements are achieved by introducing a new objective function that aims to minimize spatial as well as temporal reconstruction losses simultaneously. These losses are defined using bi-linear sampling kernel and penalized using the Charbonnier penalty function. The objective function, thus created, provides robustness to image gradient noises thereby improving the overall estimation accuracy without resorting to any coarse to fine strategies which are currently prevalent in the literature. Another novelty lies in the fact that we combine a disparity-based depth estimation network with a pose estimation network to obtain absolute scale-aware 6 DOF Camera pose and superior depth map. The effectiveness of the proposed approach is demonstrated through performance comparison with the existing supervised and unsupervised methods on the KITTI driving dataset.

AB - This paper presents an unsupervised deep learning framework called UnDEMoN for estimating dense depth map and 6-DoF camera pose information directly from monocular images. The proposed network is trained using unlabeled monocular stereo image pairs and is shown to provide superior performance in depth and ego-motion estimation compared to the existing state-of-the-art. These improvements are achieved by introducing a new objective function that aims to minimize spatial as well as temporal reconstruction losses simultaneously. These losses are defined using bi-linear sampling kernel and penalized using the Charbonnier penalty function. The objective function, thus created, provides robustness to image gradient noises thereby improving the overall estimation accuracy without resorting to any coarse to fine strategies which are currently prevalent in the literature. Another novelty lies in the fact that we combine a disparity-based depth estimation network with a pose estimation network to obtain absolute scale-aware 6 DOF Camera pose and superior depth map. The effectiveness of the proposed approach is demonstrated through performance comparison with the existing supervised and unsupervised methods on the KITTI driving dataset.

UR - http://www.mendeley.com/research/undemon-unsupervised-deep-network-depth-egomotion-estimation

U2 - 10.1109/IROS.2018.8593864

DO - 10.1109/IROS.2018.8593864

M3 - Chapter

SN - 9781538680940

T3 - IEEE International Conference on Intelligent Robots and Systems

SP - 1082

EP - 1088

BT - IEEE International Conference on Intelligent Robots and Systems

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Madhu Babu V, Das K, Majumdar A, Kumar S. UnDEMoN: Unsupervised Deep Network for Depth and Ego-Motion Estimation. In IEEE International Conference on Intelligent Robots and Systems. Institute of Electrical and Electronics Engineers Inc. 2018. p. 1082-1088. (IEEE International Conference on Intelligent Robots and Systems). https://doi.org/10.1109/IROS.2018.8593864