Final Report (Final Report)

Download File

M21_Cover_Page__1.docx.pdf

Description

A high-quality driving dataset is a key infrastructure to thrive in the autonomous vehicle industry in Pittsburgh and build a smart city for the residence. In this project, we aim to build the world’s first scenario-based driving database that is dedicated to connected and autonomous vehicles. We plan to record and model the dynamic traffic information in Pittsburgh from heterogeneous driving data such as lidar point cloud, vision information, GPS, etc. Then, scalable machine learning approaches, including unsupervised learning, will be applied to automatically facilitate the extraction of typical driving scenarios. Our data collection platform is equipped with multiple advanced sensors including Lidar, high-resolution camera, radar, GPS, IMU units, and vehicle information such as steering wheels and braking pedals. The platform is able to capture the complex and informative real-world driving scenarios and categorize them as high-dimensional and heterogeneous time series data. After that, we first propose several representative vehicle behavior categories and traffic scenarios in order to extract semantic segmentation from traffic data. An unsupervised learning approach based on nonparametric Bayesian will also be applied to learn and recognize driving scenarios. A user-friendly web application will be developed to provide the dataset to the public from a scenario perspective.

Citation

Publication Date

Jan. 2, 2020