Communication-Efficient Cooperative SLAMMOT via Determining the Number of Collaboration Vehicles

TL;DR

This work proposes a LiDAR-based communication-efficient C-SLAMMOT (CE C-SLAMMOT) method that achieves a good trade-off between performance and communication costs, while also outperforms previous state-of-the-art methods in cooperative perception performance.

Abstract

The SLAMMOT, i.e. simultaneous localization, mapping, and moving object (detection and) tracking, represents an emerging technology for autonomous vehicles in dynamic environments. Such single-vehicle systems still have inherent limitations, such as occlusion issues. Inspired by SLAMMOT and rapidly evolving cooperative technologies, it is natural to explore cooperative simultaneous localization, mapping, moving object (detection and) tracking (C-SLAMMOT) to enhance state estimation for ego-vehicles and moving objects. C-SLAMMOT could significantly upgrade the single-vehicle performance by utilizing and integrating the shared information through communication among the multiple vehicles. This inevitably leads to a fundamental trade-off between performance and communication cost, especially in a scalable manner as the number of collaboration vehicles increases. To address this challenge, we propose a LiDAR-based communication-efficient C-SLAMMOT (CE C-SLAMMOT) method by determining the number of collaboration vehicles. In CE C-SLAMMOT, we adopt descriptor-based methods for enhancing ego-vehicle pose estimation and spatial confidence map-based methods for cooperative object perception, allowing for the continuous and dynamic selection of the corresponding critical collaboration vehicles and interaction content. This approach avoids the waste of precious communication costs by preventing the sharing of information from certain collaborative vehicles that may contribute little or no performance gain, compared to the baseline method of exchanging raw observation information among all vehicles. Comparative experiments in various aspects have confirmed that the proposed method achieves a good trade-off between performance and communication costs, while also outperforms previous state-of-the-art methods in cooperative perception performance.

Figures (5)

• Figure 1: Architecture of the presented CE C-SLAMMOT.
• Figure 2: The factor graph model in proposed CE C-SLAMMOT solution. (a) Joint factor graph optimization for state estimation of ego-vehicle and moving objects. (b) is a detail subfigure of inter-vehicle data association factor. (c) is an explanation subfigure of asynchronous object state estimation.
• Figure 3: Ego-trajectory error maps for presented method with different numbers of collaboration vehicles in cooperative SLAM module. (a) and (b) are the results in curved-road scene and intersection scene of OPV2V. The gray dashed line denotes the ground truth.
• Figure 4: The results of CE CoBEVT selecting cooperative vehicles. (a) and (b) denote the number of vehicles dynamically selected (including the ego-vehicle) for participation in cooperative perception under different detection precision (AP@0.5) in curved-road and intersection scenes, respectively. The black line denotes the total number of vehicles participating in cooperative perception without selection (not always four due to maximum communication distance limitations) with the corresponding detection precision. The blue, green, and red lines denote the changes in the number of cooperative vehicles with the corresponding detection precision under different selection situations.
• Figure 5: The visualization results of the proposed CE C-SLAMMOT in the curved-road (a) of OPV2V and the cross-intersection (b) of V2V4Real. (c) and (d) are the comparison of the estimated trajectories with different methods for object 9 in the curved-road and object 15 in the cross-intersection, respectively. The red, blue, green and olive rectangular solid denote the ego-vehicle and neighboring vehicles. The hollow green bounding box and midnight-blue rectangular solid denote detected and tracked objects with IDs. The green solid line denotes the pose estimation of the cooperative vehicles, while the dashed lines in different colors denote the tracking trajectories.