Predicting Driver's Perceived Risk: a Model Based on Semi-Supervised Learning Strategy

2025-04-17

AI Summary

To address the challenge of accurately modeling drivers’ subjective perception of risk in autonomous driving, this paper proposes a Dynamic Heterogeneous Decay Risk Model—the first to characterize perceived risk as a coupled process involving dynamic triggering, anisotropic spatial sensitivity, and temporal decay. Leveraging real-time subjective risk ratings collected from driver-in-the-loop experiments, we design a CNN-Bi-LSTM-TPA hybrid network and integrate a semi-supervised learning strategy to mitigate noise inherent in subjective annotations. The approach achieves remarkable robustness under limited-data conditions, attaining a prediction accuracy of 87.91%—a 20.12 percentage-point improvement over the state-of-the-art. It also outperforms four major LSTM variants. This work establishes a novel, quantifiable paradigm for subjective risk modeling, advancing safety assessment of autonomous driving systems and optimizing human–machine trust calibration.

Technology Category

• Natural Language Processing: Safety and Robustness
• Reasoning under Uncertainty: Sequential Decision Making
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Drivers' perception of risk determines their acceptance, trust, and use of the Automated Driving Systems (ADSs). However, perceived risk is subjective and difficult to evaluate using existing methods. To address this issue, a driver's subjective perceived risk (DSPR) model is proposed, regarding perceived risk as a dynamically triggered mechanism with anisotropy and attenuation. 20 participants are recruited for a driver-in-the-loop experiment to report their real-time subjective risk ratings (SRRs) when experiencing various automatic driving scenarios. A convolutional neural network and bidirectional long short-term memory network with temporal pattern attention (CNN-Bi-LSTM-TPA) is embedded into a semi-supervised learning strategy to predict SRRs, aiming to reduce data noise caused by subjective randomness of participants. The results illustrate that DSPR achieves the highest prediction accuracy of 87.91% in predicting SRRs, compared to three state-of-the-art risk models. The semi-supervised strategy improves accuracy by 20.12%. Besides, CNN-Bi-LSTM-TPA network presents the highest accuracy among four different LSTM structures. This study offers an effective method for assessing driver's perceived risk, providing support for the safety enhancement of ADS and driver's trust improvement.

Problem

• Predicting subjective driver risk perception in automated driving systems
• Reducing data noise from subjective risk ratings using semi-supervised learning
• Improving accuracy of risk models for autonomous vehicle safety enhancement

Innovation

• Semi-supervised learning reduces subjective data noise
• CNN-Bi-LSTM-TPA network enhances prediction accuracy
• Dynamic anisotropic risk model improves ADS safety