Cooperative Perception for 3D Object Detection in Driving Scenarios Using Infrastructure Sensors

Abstract

Please refer to published version for the most recent bibliographic citation information. If a published version is known of, the repository item page linked to above, will contain details on accessing it.

AutoC2X: Open-source software to realize V2X cooperative perception among autonomous vehicles

“…Some researchers also evaluated the realization of featurelevel sensor data sharing in real vehicles [15]. Furthermore, object(track)-level cooperative perception has been realized using infrastructure involving multiple sensors and multiple transmitters [16], [17]. In [18], the researchers attempted to examine the trustworthiness of cooperative perception by quantifying the confidence in the correctness of data by using the Bayes theory.…”

Fast and Robust Registration of Partially Overlapping Point Clouds

“…This is a key problem in many downstream applications including 3D scene reconstruction [1], localisation [2] and SLAM [3]. Recent applications such as Augmented Reality (AR) [4], cooperative (multi-agent) perception for autonomous vehicles [5] and multi-agent SLAM [6] introduce new challenges to this problem. Specifically, these applications require registration methods that are robust to point clouds with low overlap, e.g.…”

“…SLAM [3] and lidar odometry [7], where pairs of point clouds are obtained sequentially in adjacent time steps by a single vehicle navigating in a driving environment. On the other hand, applications such as cooperative perception [5] and multi-agent SLAM [6] require registering point clouds obtained simultaneously from a pair of sensors on two different vehicles that are potentially far apart, and thus, may have low field-of-view overlap, e.g. Figure 1.…”

Research on Vehicle Automatic Driving Target Perception Technology Based on Improved MSRPN Algorithm

“…However, the algorithm is still rough in target detection and classification. [10][11][12] On this basis, this research has been optimized and improved, hoping to propose a vehicle automatic driving target recognition algorithm with excellent performance.…”