ALERT Open Dataset and Input-Size-Agnostic Vision Transformer for Driver Activity Recognition using IR-UWB

2025-12-13

AI Summary

This work addresses two key bottlenecks in distracted driving recognition: (1) the absence of large-scale, real-world impulse radio ultra-wideband (IR-UWB) radar datasets, and (2) the incompatibility of standard Vision Transformers (ViTs) with irregular UWB signal dimensions. To this end, we introduce ALERT—the first large-scale, real-driving-scenario IR-UWB dataset—comprising 10,220 samples across seven driver behavior classes. We further propose ISA-ViT, an input-size-agnostic ViT architecture featuring adaptive patch partitioning and transferable pretrained positional encoding to preserve Doppler-phase time-frequency features, along with a dedicated time-frequency feature fusion mechanism. On the UWB-based distracted driving recognition (UWB-DAR) task, ISA-ViT achieves a 22.68% absolute accuracy improvement over the baseline ViT. The ALERT dataset is publicly released, establishing foundational data and model resources for privacy-preserving in-vehicle perception.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Planning, Routing, and Scheduling: Activity and Plan Recognition
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Data and user privacy-enhancing technologies for the Web
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

Distracted driving contributes to fatal crashes worldwide. To address this, researchers are using driver activity recognition (DAR) with impulse radio ultra-wideband (IR-UWB) radar, which offers advantages such as interference resistance, low power consumption, and privacy preservation. However, two challenges limit its adoption: the lack of large-scale real-world UWB datasets covering diverse distracted driving behaviors, and the difficulty of adapting fixed-input Vision Transformers (ViTs) to UWB radar data with non-standard dimensions. This work addresses both challenges. We present the ALERT dataset, which contains 10,220 radar samples of seven distracted driving activities collected in real driving conditions. We also propose the input-size-agnostic Vision Transformer (ISA-ViT), a framework designed for radar-based DAR. The proposed method resizes UWB data to meet ViT input requirements while preserving radar-specific information such as Doppler shifts and phase characteristics. By adjusting patch configurations and leveraging pre-trained positional embedding vectors (PEVs), ISA-ViT overcomes the limitations of naive resizing approaches. In addition, a domain fusion strategy combines range- and frequency-domain features to further improve classification performance. Comprehensive experiments demonstrate that ISA-ViT achieves a 22.68% accuracy improvement over an existing ViT-based approach for UWB-based DAR. By publicly releasing the ALERT dataset and detailing our input-size-agnostic strategy, this work facilitates the development of more robust and scalable distracted driving detection systems for real-world deployment.

Problem

• Lack of large-scale real-world UWB datasets for diverse distracted driving behaviors
• Difficulty adapting fixed-input Vision Transformers to non-standard UWB radar data dimensions
• Need to preserve radar-specific information like Doppler shifts during data processing

Innovation

• Introduces ALERT dataset with 10,220 real-world IR-UWB radar samples
• Proposes input-size-agnostic Vision Transformer for non-standard radar data
• Uses domain fusion of range and frequency features to boost accuracy