Sign in Subscribe
Machine learning

Emerging Technologies: Researchers Develop Automated System to Detect Cellphone Use During Driving

Strategic Insights

— 1 min read

Researchers from K.N. Toosi University of Technology in Tehran, Iran, have developed an automated system using machine learning and computer vision to detect cellphone use during driving. The system, which uses Wavelet Scattering Networks and Convolutional Neural Networks, achieved a test accuracy of 91% for a new dataset and 99% for a publicly available dataset. This breakthrough has significant implications for road safety, particularly in regions with high rates of distracted driving.

Key Takeaways:

  • The automated system uses machine learning and computer vision to detect cellphone use during driving, with a focus on monitoring the driver's face and body from two viewpoints.
  • The system uses Wavelet Scattering Networks, a simple and efficient type of architecture, and requires little hyper-parameter tuning.
  • The speed of this model is similar to Convolutional Neural Networks, making it a viable option for real-time applications.
  • The system achieved a test accuracy of 91% for a new dataset and 99% for a publicly available dataset, demonstrating its effectiveness in detecting cellphone use during driving.
  • The researchers used a combination of datasets, including a new dataset created specifically for the study and a publicly available dataset.
  • The study highlights the importance of emerging technologies in improving road safety, particularly in regions with high rates of distracted driving.

Statistics:

  • The system achieved a test accuracy of 91% for a new dataset created specifically for the study.
  • The system achieved a test accuracy of 99% for a publicly available dataset used in the study.
  • The system uses a combination of machine learning and computer vision to detect cellphone use during driving, with a focus on monitoring the driver's face and body from two viewpoints.

Sources:

  • "Driver cellphone usage detection using wavelet scattering and convolutional neural networks" (AUT Journal of Mathematics and Computing, 2025, Article 6(3):257-268)
  • "K.N. Toosi University of Technology Researchers Detail Findings in Emerging Technologies (Driver cellphone usage detection using wavelet scattering and convolutional neural networks)" (Journal of Engineering, July 7, 2025)

Read more