New AI model estimates car speed using only smartphone accelerometer data

Researchers have developed CarSpeedNet, a novel machine learning model capable of estimating vehicle speed using only accelerometer data from a smartphone. This model does not require gyroscope, GPS, or vehicle-specific data during inference. Experiments showed that a 4-second input window achieved a root-mean-square error of 1.8 m/s, significantly outperforming a 1-second window which resulted in a 2.9 m/s error. AI

IMPACT This model could enable new applications in vehicle analytics and driver behavior monitoring without requiring complex sensor setups.

RANK_REASON Publication of a research paper detailing a new machine learning model. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Barak Or
• CarSpeedNet

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →