Hierarchical Spatio-Temporal Attention Network with Adaptive Risk-Aware Decision for Forward Collision Warning in Complex Scenarios

TL;DR

This work tackles FCW reliability in complex scenarios by decoupling perception-prediction from adaptive decision-making. It introduces HSTAN, a lightweight Hierarchical Spatio-Temporal Attention Network with SAM (GAT-based) and TAM (GRU plus self-attention) to predict multi-agent trajectories and quantify uncertainty via Conformalized Quantile Regression. The Dynamic Risk Threshold Adjustment (DRTA) module combines a physics-informed risk potential with an adaptive, SPC-inspired threshold to convert predictions into timely, reliable warnings, achieving state-of-the-art F1 (0.912) and low false alarms (8.2%) while maintaining substantial lead times (2.8 s). Across diverse datasets and scenarios, the approach demonstrates strong accuracy, efficiency suitable for edge deployment, and robust warning reliability, signaling practical impact for advanced FCW in real-world intelligent transportation systems.

Abstract

Forward Collision Warning systems are crucial for vehicle safety and autonomous driving, yet current methods often fail to balance precise multi-agent interaction modeling with real-time decision adaptability, evidenced by the high computational cost for edge deployment and the unreliability stemming from simplified interaction models.To overcome these dual challenges-computational complexity and modeling insufficiency-along with the high false alarm rates of traditional static-threshold warnings, this paper introduces an integrated FCW framework that pairs a Hierarchical Spatio-Temporal Attention Network with a Dynamic Risk Threshold Adjustment algorithm. HSTAN employs a decoupled architecture (Graph Attention Network for spatial, cascaded GRU with self-attention for temporal) to achieve superior performance and efficiency, requiring only 12.3 ms inference time (73% faster than Transformer methods) and reducing the Average Displacement Error (ADE) to 0.73m (42.2% better than Social_LSTM) on the NGSIM dataset. Furthermore, Conformalized Quantile Regression enhances reliability by generating prediction intervals (91.3% coverage at 90% confidence), which the DTRA module then converts into timely warnings via a physics-informed risk potential function and an adaptive threshold mechanism inspired by statistical process control.Tested across multi-scenario datasets, the complete system demonstrates high efficacy, achieving an F1 score of 0.912, a low false alarm rate of 8.2%, and an ample warning lead time of 2.8 seconds, validating the framework's superior performance and practical deployment feasibility in complex environments.

Figures (13)

• Figure 1: Overall Framework
• Figure 2: HSTAN Architecture Diagram
• Figure 3: GAT Architecture Diagram
• Figure 4: MHA Architecture Diagram
• Figure 5: DTRA Architecture Diagram