Towards Collaborative Perception for Automated Vehicles in Heterogeneous Traffic

Khan, Saifullah; Andert, Franz; Wojke, Nicolai; Schindler, Julian; Correa, Alejandro; Wijbenga, Anton

doi:10.1007/978-3-319-99762-9_3

Cited by 9 publications

(5 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the impact of the vehicle awareness ratio on the vehicle capability to correctly assess its collision risk based on its environmental model has not been assessed to date. Collective perception is investigated in several European collaborative research projects, such as TransAID [ 15 ], AutoNet2030 [ 16 ], IMAGinE [ 17 ] and MAVEN [ 18 ].…”

Section: Collective Perceptionmentioning

confidence: 99%

Collective Perception: A Safety Perspective

Schiegg

Llátser

Bischoff³

et al. 2020

Sensors

View full text Add to dashboard Cite

Vehicle-to-everything (V2X) communication is seen as one of the main enabling technologies for automated vehicles. Collective perception is especially promising, as it allows connected traffic participants to “see through the eyes of others” by sharing sensor-detected objects via V2X communication. Its benefit is typically assessed in terms of the increased object update rate, redundancy, and awareness. To determine the safety improvement thanks to collective perception, the authors introduce new metrics, which quantify the environmental risk awareness of the traffic participants. The performance of the V2X service is then analyzed with the help of the test platform TEPLITS, using real traffic traces from German highways, amounting to over 100 h of total driving time. The results in the considered scenarios clearly show that collective perception not only contributes to the accuracy and integrity of the vehicles’ environmental perception, but also that a V2X market penetration of at least 25% is necessary to increase traffic safety from a “risk of serious traffic accidents” to a “residual hypothetical risk of collisions without minor injuries” for traffic participants equipped with non-redundant 360° sensor systems. These results support the ongoing worldwide standardization efforts of the collective perception service.

show abstract

Section: Collective Perceptionmentioning

confidence: 99%

Collective Perception: A Safety Perspective

Schiegg

Llátser

Bischoff³

et al. 2020

Sensors

View full text Add to dashboard Cite

show abstract

“…For example, Comm-Net [32] adopted the averaging operation to achieve continuous communication in multi-agent system; VAIN [11] considered an attention mechanism to determine which agents would share information; Collaborative perception. Collaborative perception is an application of multi-agent communication system to perception tasks [9,36,14]. Who2com [20] exploited a handshake communication mechanism to determine which two agents should communicate for image segmentation; When2com [19] introduced an asymmetric attention mechanism to decide when to communicate and how to create communication groups for image segmentation; V2VNet [34] proposed multiple rounds of message passing on a spatial-aware graph neural network for joint perception and prediction in autonomous driving; and [33] proposed a pose error regression module to learn to correct pose errors when the pose information from other agents is noisy.…”

Section: Related Workmentioning

confidence: 99%

Learning Distilled Collaboration Graph for Multi-Agent Perception

Ren²,

et al. 2021

Preprint

View full text Add to dashboard Cite

To promote better performance-bandwidth trade-off for multi-agent perception, we propose a novel distilled collaboration graph (DiscoGraph) to model trainable, pose-aware, and adaptive collaboration among agents. Our key novelties lie in two aspects. First, we propose a teacher-student framework to train DiscoGraph via knowledge distillation. The teacher model employs an early collaboration with holistic-view inputs; the student model is based on intermediate collaboration with single-view inputs. Our framework trains DiscoGraph by constraining postcollaboration feature maps in the student model to match the correspondences in the teacher model. Second, we propose a matrix-valued edge weight in DiscoGraph. In such a matrix, each element reflects the inter-agent attention at a specific spatial region, allowing an agent to adaptively highlight the informative regions. During inference, we only need to use the student model named as the distilled collaboration network (DiscoNet). Attributed to the teacher-student framework, multiple agents with the shared DiscoNet could collaboratively approach the performance of a hypothetical teacher model with a holistic view. Our approach is validated on V2X-Sim 1.0, a large-scale multi-agent perception dataset that we synthesized using CARLA and SUMO co-simulation. Our quantitative and qualitative experiments in multi-agent 3D object detection show that DiscoNet could not only achieve a better performance-bandwidth trade-off than the state-of-the-art collaborative perception methods, but also bring more straightforward design rationale. Our code is available on https://github.com/ai4ce/DiscoNet.

show abstract

“…Infrastructure-assisted CP is also part of the scope of Managing Automated Vehicles Enhances Network (MAVEN) [ 45 ], an EU funded project targeting traffic management solutions where CAVs are guided at signalised cooperative intersections in urban traffic environments [ 40 ]. Other CP related joint research projects include TransAID [ 46 ] and IMAGinE [ 47 ].…”

Section: Related Workmentioning

confidence: 99%

Demonstrations of Cooperative Perception: Safety and Robustness in Connected and Automated Vehicle Operations

Shan

Narula

Wong³

et al. 2020

Sensors

View full text Add to dashboard Cite

Cooperative perception, or collective perception (CP), is an emerging and promising technology for intelligent transportation systems (ITS). It enables an ITS station (ITS-S) to share its local perception information with others by means of vehicle-to-X (V2X) communication, thereby achieving improved efficiency and safety in road transportation. In this paper, we present our recent progress on the development of a connected and automated vehicle (CAV) and intelligent roadside unit (IRSU). The main contribution of the work lies in investigating and demonstrating the use of CP service within intelligent infrastructure to improve awareness of vulnerable road users (VRU) and thus safety for CAVs in various traffic scenarios. We demonstrate in experiments that a connected vehicle (CV) can “see” a pedestrian around the corners. More importantly, we demonstrate how CAVs can autonomously and safely interact with walking and running pedestrians, relying only on the CP information from the IRSU through vehicle-to-infrastructure (V2I) communication. This is one of the first demonstrations of urban vehicle automation using only CP information. We also address in the paper the handling of collective perception messages (CPMs) received from the IRSU, and passing them through a pipeline of CP information coordinate transformation with uncertainty, multiple road user tracking, and eventually path planning/decision-making within the CAV. The experimental results were obtained with manually driven CV, fully autonomous CAV, and an IRSU retrofitted with vision and laser sensors and a road user tracking system.

show abstract

Towards Collaborative Perception for Automated Vehicles in Heterogeneous Traffic

Cited by 9 publications

References 17 publications

Collective Perception: A Safety Perspective

Collective Perception: A Safety Perspective

Learning Distilled Collaboration Graph for Multi-Agent Perception

Demonstrations of Cooperative Perception: Safety and Robustness in Connected and Automated Vehicle Operations

Contact Info

Product

Resources

About