Abstract:Surrounding perceptions are quintessential for safe driving for connected and autonomous vehicles (CAVs), where the Bird's Eye View has been employed to accurately capture spatial relationships among vehicles. However, severe inherent limitations of BEV, like blind spots, have been identified. Collaborative perception has emerged as an effective solution to overcoming these limitations through data fusion from multiple views of surrounding vehicles. While most existing collaborative perception strategies adopt a fully connected graph predicated on fairness in transmissions, they often neglect the varying importance of individual vehicles due to channel variations and perception redundancy. To address these challenges, we propose a novel Priority-Aware Collaborative Perception (PACP) framework to employ a BEV-match mechanism to determine the priority levels based on the correlation between nearby CAVs and the ego vehicle for perception. By leveraging submodular optimization, we find near-optimal transmission rates, link connectivity, and compression metrics. Moreover, we deploy a deep learning-based adaptive autoencoder to modulate the image reconstruction quality under dynamic channel conditions. Finally, we conduct extensive studies and demonstrate that our scheme significantly outperforms the state-of-the-art schemes by 8.27\% and 13.60\%, respectively, in terms of utility and precision of the Intersection over Union.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to improve the efficiency and accuracy of connected and autonomous vehicles (CAVs) in collaborative perception while reducing the consumption of communication resources. Specifically, the paper proposes solutions to the following challenges: 1. **Blind - spot problem**: Traditional single - vehicle systems have blind areas that cannot be observed in high - density traffic scenarios, which limits their perception capabilities. 2. **Data redundancy**: Existing fairness schemes may lead to multiple vehicles transmitting duplicate data, resulting in bandwidth waste. 3. **Lack of priority**: Existing fairness schemes cannot prioritize data transmission according to the importance of vehicles (such as distance and channel conditions), which may block the transmission of more critical information. 4. **Insufficient dynamic adaptability**: Existing methods often assume ideal communication channels and ignore the impact of network capacity fluctuations in the actual environment. To address these challenges, the paper proposes a new **Priority - Aware Framework (PACP)**, which improves collaborative perception in the following aspects: - **BEV matching mechanism**: By calculating the intersection - over - union (IoU) of BEV features between adjacent vehicles, the priority weights of each vehicle are determined, and the vehicle with the most consistent perception result with the ego - vehicle is selected for data fusion. - **Sub - modular optimization**: Using sub - modular optimization techniques, jointly optimize the transmission rate, link connectivity and compression ratio to achieve approximately optimal resource allocation. - **Adaptive compression**: Introduce a deep - learning - based adaptive auto - encoder to adjust the image reconstruction quality according to dynamic channel conditions and improve data transmission efficiency. Through these innovations, the PACP framework proposed in the paper shows significantly better performance than existing methods in experiments, especially with an 8.27% increase in the utility value and a 13.60% increase in the average precision of intersection - over - union (AP@IoU).

PACP: Priority-Aware Collaborative Perception for Connected and Autonomous Vehicles

Slim-FCP: Lightweight-Feature-Based Cooperative Perception for Connected Automated Vehicles

Infrastructure-Assisted Collaborative Perception in Automated Valet Parking: A Safety Perspective

Efficient Vehicular Collaborative Perception Based on Saptial-Temporal Feature Compression

Toward Collaborative Occlusion-free Perception in Connected Autonomous Vehicles

Collaborative Perception for Connected and Autonomous Driving: Challenges, Possible Solutions and Opportunities

Direct-CP: Directed Collaborative Perception for Connected and Autonomous Vehicles via Proactive Attention

RSU-Aided Energy-Efficient Collaborative Perception for Connected Autonomous Vehicles

EMP: edge-assisted multi-vehicle perception

Decentralized Multiagent Reinforcement Learning-based Cooperative Perception with Dual-functional Radar-Communication V2V Links

Adaptive Communications in Collaborative Perception with Domain Alignment for Autonomous Driving

CP-Guard: Malicious Agent Detection and Defense in Collaborative Bird's Eye View Perception

Enhanced Cooperative Perception for Autonomous Vehicles Using Imperfect Communication

Low-Latency Perception Sharing Services for Connected Autonomous Vehicles

Model-Assisted Learning for Adaptive Cooperative Perception of Connected Autonomous Vehicles

V2VFormer++: Multi-Modal Vehicle-to-Vehicle Cooperative Perception Via Global-Local Transformer

V2V Based Visual Cooperative Perception for Connected Autonomous Vehicles: Far-Sight and See-Through

Collaborative Joint Perception and Prediction for Autonomous Driving

CoBEVFusion: Cooperative Perception with LiDAR-Camera Bird's-Eye View Fusion

EdgeCooper: Network-Aware Cooperative LiDAR Perception for Enhanced Vehicular Awareness

Vision beyond the Field-of-View: A Collaborative Perception System to Improve Safety of Intelligent Cyber-Physical Systems