Resource-Aware Algorithms for Distributed Loop Closure Detection with Provable Performance Guarantees

Tian, Yulun; Khosoussi, Kasra; How, Jonathan P.

doi:10.1007/978-3-030-44051-0_25

Cited by 6 publications

(12 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Both theorems in this section then follow as special cases. While these results are established independently (Tian et al, 2018b), we note that an even more general case with an arbitrary matroid constraint is presented in the seminal work of Nemhauser et al (1978: Proposition 3.1). Theorem 5. Let E x = E 1 ⊎ ⋯ ⊎ E n b .…”

Section: A1 Proofs Of Theorems 1 Andmentioning

confidence: 82%

“…Similar computational budgets were considered previously (Carlone and Karaman, 2017; Khosoussi et al, 2019, 2016b) for selecting informative measurements in SLAM. The present work extends Tian et al (2018b) and provides approximation algorithms with performance guarantees for inter-robot loop-closure detection under both communication and computational budgets.…”

Section: Related Workmentioning

confidence: 95%

“…In an early version of the this work, we extended the formulation in Tian et al (2018a) to scenarios where, in addition to a communication budget, robots are also subject to a computational budget (Tian et al, 2018b). Computational budget puts a limit on the number of geometric verifications and, subsequently, the number of loop closures (NLC) that will be added to the pose graph, which allows us to control the cost of CSLAM back-end.…”

Section: Related Workmentioning

confidence: 99%

“…In this section, we study Problem 1 when f : 2 E x → R ≥ 0 is an arbitrary NMS objective. To the best of the authors’ knowledge, no prior work exists on approximation algorithms for this problem except (Tian et al, 2018b). The approach we took in Section 4 for the special case of monotone modular objectives cannot be extended to this case: even evaluating the “generalized” g (8) is NP-hard in general.…”

Section: Algorithms For Submodular Performance Metricsmentioning

confidence: 99%

“…An early version of this work was presented at WAFR 2018 (Workshop on the Algorithmic Foundations of Robotics) (Tian et al, 2018b). This article improves Tian et al (2018b) and our earlier work in Tian et al (2018a) along the following directions.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A resource-aware approach to collaborative loop-closure detection with provable performance guarantees

Tian

Khosoussi

How

2020

The International Journal of Robotics Research

Self Cite

View full text Add to dashboard Cite

This paper presents resource-aware algorithms for distributed inter-robot loop-closure detection for applications such as collaborative simultaneous localization and mapping (CSLAM) and distributed image retrieval. In real-world scenarios, this process is resource-intensive as it involves exchanging many observations and geometrically verifying a large number of potential matches. This poses severe challenges for small-size and low-cost robots with various operational and resource constraints that limit, e.g., energy consumption, communication bandwidth, and computation capacity. This paper proposes a framework in which robots first exchange compact queries to identify a set of potential loop closures. We then seek to select a subset of potential inter-robot loop closures for geometric verification that maximizes a monotone submodular performance metric without exceeding budgets on computation (number of geometric verifications) and communication (amount of exchanged data for geometric verification). We demonstrate that this problem is, in general, NP-hard, and present efficient approximation algorithms with provable a priori performance guarantees. The proposed framework is extensively evaluated on real and synthetic datasets. A natural convex relaxation scheme is also presented to certify the near-optimal performance of the proposed framework a posteriori.

show abstract

Section: A1 Proofs Of Theorems 1 Andmentioning

confidence: 82%

Section: Related Workmentioning

confidence: 95%

Section: Related Workmentioning

confidence: 99%

Section: Algorithms For Submodular Performance Metricsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A resource-aware approach to collaborative loop-closure detection with provable performance guarantees

Tian

Khosoussi

How

2020

The International Journal of Robotics Research

Self Cite

View full text Add to dashboard Cite

show abstract

Enhancing Privacy in Robotics via Judicious Sensor Selection

Eick

Antón

2020

2020 IEEE International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

Communication Planning for Cooperative Terrain-Based Underwater Localization

Anderson

Hollinger

2021

Sensors

View full text Add to dashboard Cite

This paper presents a decentralized communication planning algorithm for cooperative terrain-based navigation (dec-TBN) with autonomous underwater vehicles. The proposed algorithm uses forward simulation to approximate the value of communicating at each time step. The simulations are used to build a directed acyclic graph that can be searched to provide a minimum cost communication schedule. Simulations and field trials are used to validate the algorithm. The simulations use a real-world bathymetry map from Lake Nighthorse, CO, and a sensor model derived from an Ocean Server Iver2 vehicle. The simulation results show that the algorithm finds a communication schedule that reduces communication bandwidth by 86% and improves robot localization by up to 27% compared to non-cooperative terrain-based navigation. Field trials were conducted in Foster Reservoir, OR, using two Riptide Autonomous Solutions micro-unmanned underwater vehicles. The vehicles collected GPS, altimeter, acoustic communications, and dead reckoning data while following paths on the surface of the reservoir. The data were used to evaluate the planning algorithm. In three of four missions, the planning algorithm improved dec-TBN localization while reducing acoustic communication bandwidth by 56%. In the fourth mission, dec-TBN performed better when using full communications bandwidth, but the communication policy for that mission maintained 86% of the localization accuracy while using 9% of the communications. These results indicate that the presented communication planning algorithm can maintain or improve dec-TBN accuracy while reducing the number of communications used for localization.

show abstract

Resource-Aware Algorithms for Distributed Loop Closure Detection with Provable Performance Guarantees

Cited by 6 publications

References 27 publications

A resource-aware approach to collaborative loop-closure detection with provable performance guarantees

A resource-aware approach to collaborative loop-closure detection with provable performance guarantees

Enhancing Privacy in Robotics via Judicious Sensor Selection

Communication Planning for Cooperative Terrain-Based Underwater Localization

Contact Info

Product

Resources

About