CarSpeedNet: A Deep Neural Network-based Car Speed Estimation from Smartphone Accelerometer
This provides a method for car speed estimation without car connectivity, which is incremental for applications like traffic monitoring or driver assistance.
The paper tackled the problem of estimating car speed using only smartphone accelerometer data, achieving a precision of less than 0.72 m/s in extended driving tests.
We introduce the CarSpeedNet, a deep learning model designed to estimate car speed using three-axis accelerometer data from smartphones. Using 13 hours of data collected from a smartphone in cars across various roads, CarSpeedNet accurately models the relationship between smartphone acceleration and car speed. Ground truth speed data was collected at 1 [Hz] from GPS receivers. The model provides high-frequency speed estimation by incorporating historical data and achieves a precision of less than 0.72 [m/s] during extended driving tests, relying solely on smartphone accelerometer data without any connection to the car.