2020
DOI: 10.3390/s20133737
|View full text |Cite
|
Sign up to set email alerts
|

Joint Unsupervised Learning of Depth, Pose, Ground Normal Vector and Ground Segmentation by a Monocular Camera Sensor

Abstract: We propose a completely unsupervised approach to simultaneously estimate scene depth, ego-pose, ground segmentation and ground normal vector from only monocular RGB video sequences. In our approach, estimation for different scene structures can mutually benefit each other by the joint optimization. Specifically, we use the mutual information loss to pre-train the ground segmentation network and before adding the corresponding self-learning label obtained by a geometric method. By using the static natur… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
10
0

Year Published

2021
2021
2024
2024

Publication Types

Select...
3
2

Relationship

0
5

Authors

Journals

citations
Cited by 5 publications
(10 citation statements)
references
References 38 publications
0
10
0
Order By: Relevance
“…The angle between predicted and ground-truth normal is used as the evaluation metric. From Table . 3 we can see that the mean error of our method is 1.12 degree which are significantly smaller than other unsupervised methods [8] [42].…”
Section: Ground Plane Normal Evaluationmentioning
confidence: 61%
See 2 more Smart Citations
“…The angle between predicted and ground-truth normal is used as the evaluation metric. From Table . 3 we can see that the mean error of our method is 1.12 degree which are significantly smaller than other unsupervised methods [8] [42].…”
Section: Ground Plane Normal Evaluationmentioning
confidence: 61%
“…Explicit methods [18,42] usually translate the homography estimation problem into two sub-problems: ego-motion and ground plane estimation, assuming the camera intrinsic parameters are known previously. Hartley et.al [18] first used the normal vectors of the coplanar points and egomotion to construct a homography.…”
Section: Homography Estimationmentioning
confidence: 99%
See 1 more Smart Citation
“…Currently, most of the work related to improving edge depth requires the introduction of an additional network, e.g., semantic segmentation [10][11][12], edge map detection networks [13][14][15], or optical flow [16]. We found that research on uncertainty, which has only recently entered the limelight, can also improve the quality of edge depth and without learning other complex networks.…”
Section: Introductionmentioning
confidence: 92%
“…e latter three have significant advantages and disadvantages. e monocular camera only needs one camera, which is low cost, small volume, and high compatibility with application scenes [17]. e monocular camera and IMU are the two most important sensors in the VINS.…”
Section: 1mentioning
confidence: 99%