Multi-object Monocular SLAM for Dynamic Environments

Nair, Gokul B.; Daga, Swapnil; Sajnani, Rahul; Ramesh, Anirudha; Ansari, Junaid; Jatavallabhula, Krishna Murthy; Krishna, K. Madhava

doi:10.1109/iv47402.2020.9304648

Cited by 14 publications

(8 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, it cannot handle the features of a car winding along a flat road. Based on the fact that most objects move on flat planes, some methods [ 14 , 46 ] reconstruct features u on the ground based on the current frame using the ground plane [ n , h] (normal and distance in the camera frame), as shown in Fig. 3 b.…”

Section: Low-level-feature-based Dynamic Slammentioning

confidence: 99%

See 1 more Smart Citation

A survey: which features are required for dynamic visual simultaneous localization and mapping?

Zheng

2021

Vis. Comput. Ind. Biomed. Art

View full text Add to dashboard Cite

In recent years, simultaneous localization and mapping in dynamic environments (dynamic SLAM) has attracted significant attention from both academia and industry. Some pioneering work on this technique has expanded the potential of robotic applications. Compared to standard SLAM under the static world assumption, dynamic SLAM divides features into static and dynamic categories and leverages each type of feature properly. Therefore, dynamic SLAM can provide more robust localization for intelligent robots that operate in complex dynamic environments. Additionally, to meet the demands of some high-level tasks, dynamic SLAM can be integrated with multiple object tracking. This article presents a survey on dynamic SLAM from the perspective of feature choices. A discussion of the advantages and disadvantages of different visual features is provided in this article.

show abstract

Section: Low-level-feature-based Dynamic Slammentioning

confidence: 99%

“…To tackle the second problem, Nair et al [ 46 ] leveraged multiple sources to obtain localizations of moving objects and maintained cyclic consistencies in a pose graph. They first used the 3D coordinates of ground points obtained by Eq.…”

Section: Low-level-feature-based Dynamic Slammentioning

confidence: 99%

A survey: which features are required for dynamic visual simultaneous localization and mapping?

Zheng

2021

Vis. Comput. Ind. Biomed. Art

View full text Add to dashboard Cite

show abstract

“…Recent Approaches: A more recent approach to the multibody SLAM problem in a monocular setting is proposed by Nair et al (Nair et al, 2020) which relies on batch-based pose-graph optimization in 6 DoF. The optimization framework used in it cannot be applied in a real-time setting.…”

Section: Related Workmentioning

confidence: 99%

“…A monocular multibody approach in SE(3) similar to Nair et al (Nair et al, 2020) but the camera nodes are fed with scale-ambiguous ORB (Mur-Artal and Tardós, 2017) initialization. We show that the optimizer itself is able to pull scale-ambiguous odometry to metric scale without relying on any prior scale correction like Sec.…”

Section: Batch Optimized Baseline In Se(3) Withmentioning

confidence: 99%

See 1 more Smart Citation

BirdSLAM: Monocular Multibody SLAM in Bird's-Eye View

Daga¹,

Nair²,

Ramesh³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

In this paper, we present BirdSLAM, a novel simultaneous localization and mapping (SLAM) system for the challenging scenario of autonomous driving platforms equipped with only a monocular camera. BirdSLAM tackles challenges faced by other monocular SLAM systems (such as scale ambiguity in monocular reconstruction, dynamic object localization, and uncertainty in feature representation) by using an orthographic (bird's-eye) view as the configuration space in which localization and mapping are performed. By assuming only the height of the ego-camera above the ground, BirdSLAM leverages single-view metrology cues to accurately localize the ego-vehicle and all other traffic participants in bird's-eye view. We demonstrate that our system outperforms prior work that uses strictly greater information, and highlight the relevance of each design decision via an ablation analysis.

show abstract