Coupled-Perception-Framework-for-Road-Condition-Assessment

Published: 10 February 2026| Version 1 | DOI: 10.17632/fd5h9s2vbc.1
Contributor:
yaoqi sun

Description

Coupled-Perception-Framework-for-Road-Condition-Assessment This study introduces an innovative approach for high-precision International Roughness Index (IRI) estimation using low-cost telemetry data, addressing the balance between cost and accuracy in road condition assessment. Project Structure data/: Contains the dataset used in this paper. optuna/: Includes all base model experiments, enhanced model and ablation study experiments, visualization code, SHAP model interpretation code, and all figures in the paper. Filter.py: The filter designed in this paper. environment.yml: Environment configuration file. Environment Configuration This project was developed on a workstation with the following specifications: OS: Windows 11 CPU: Intel Core i7-12700K GPU: NVIDIA GeForce RTX 3080 RAM: 32GB DDR4 Python: 3.13.5 Key Libraries: PyTorch, Optuna, Scikit-learn, Matplotlib, Pandas, NumPy

Files

Institutions

Categories

Transportation Engineering, Smart Transportation, Road Transportation

Licence