CrashFormer: A Multimodal Architecture to Predict the Risk of Crash
Authors:
- Amin Karimi MonsefiAMAmin Karimi Monsefi
- Pouya ShiriPSPouya Shiri
- Ahmad MohammadshiraziAMAhmad Mohammadshirazi
- Nastaran Karimi MonsefiNMNastaran Karimi Monsefi
- Ron DaviesRDRon Davies
- Sobhan MoosaviSMSobhan Moosavi
- Rajiv RamnathRRRajiv Ramnath
Preprint
Abstract
Reducing traffic accidents is a crucial global public safety concern. Accident prediction is key to improving traffic safety, enabling proactive measures to be taken before a crash occurs, and informing safety policies, regulations, and targeted interventions. Despite numerous studies on accident prediction over the past decades, many have limitations in terms of generalizability, reproducibility, or feasibility for practical use due to input data or problem formulation. To address existing shortcomings, we propose CrashFormer, a multi-modal architecture that utilizes comprehensive (but relatively easy to obtain) inputs such as the history of accidents, weather information, map images, and demographic information. The model predicts the future risk of accidents on a reasonably acceptable cadence (i.e., every six hours) for a geographical location of 5.161 square kilometers. CrashFormer is composed of five components: a sequential encoder to utilize historical accidents and weather data, an image encoder to use map imagery data, a raw data encoder to utilize demographic information, a feature fusion module for aggregating the encoded features, and a classifier that accepts the aggregated data and makes predictions accordingly. Results from extensive real-world experiments in 10 major US cities show that CrashFormer outperforms state-of-the-art sequential and non-sequential models by 1.8% in F1-score on average when using ``sparse'' input data.Keywords
All Related Versions
- Version 1, 2024-02-07, ArXiv
- Published version: , 42.
This publication has 16 references indexed in Scilit:
- Smart and collaborative industrial IoT: A federated learning and data space approachDigital Communications and Networks, 2023
- Exploring the effect of road network, demographic, and land use characteristics on teen crash frequency using geographically weighted negative binomial regressionAccident Analysis & Prevention, 2022
- An integrated methodology for real-time driving risk status prediction using naturalistic driving dataAccident Analysis & Prevention, 2021
- Real-time crash risk prediction on arterials based on LSTM-CNNAccident Analysis & Prevention, 2020
- A feature learning approach based on XGBoost for driving assessment and risk predictionAccident Analysis & Prevention, 2019
- A spatiotemporal deep learning approach for citywide short-term crash risk prediction with multi-source dataAccident Analysis & Prevention, 2019
- A Deep Learning Approach to the Citywide Traffic Accident Risk Prediction2018
- A novel variable selection method based on frequent pattern tree for real-time traffic accident risk predictionTransportation Research Part C: Emerging Technologies, 2015
- The influence of rainfall on road accidents in urban areas: A weather radar approachTravel Behaviour and Society, 2014
- The mixed effects of precipitation on traffic crashesAccident Analysis & Prevention, 2004