IJOT
Vol. 6 No. 1 (2024): International Journal on Orange Technologies (IJOT)

Using a Deep Convolutional Neural Network to Identify Vehicle Driver Activity

Shynu T (Unknown)
S. Suman Rajest (Unknown)
R. Regin (Unknown)
Steffi. R (Unknown)



Article Info

Publish Date
26 Jan 2024

Abstract

Driver actions and judgments are the most critical aspects influencing passenger safety in a vehicle. Common things that drivers do include: driving safely, talking on the phone, texting, eating, reaching behind, altering hairstyle or makeup, operating the radio, and operating mobile phones. The first four are considered typical driving duties, whereas the latter seven are considered distractions. The strategy is to take in the live feed from the dashboard camera, process the frames with pre-trained convolutional neural network (CNN) models, and then refine it with the transfer learning method to identify the driver's activities. An activity recognition system for drivers is developed utilising deep Convolutional Neural Networks in order to monitor their actions (CNN). A warning is sent to the driver based on the identified driver's actions. In addition to activating the warning lights, the system gently slows down the vehicle or forces the driver to shift to the side lane and come to a complete stop in order to protect everyone in the vehicle and the surrounding area.

Copyrights © 2024






Journal Info

Abbrev

IJOT

Publisher

Subject

Automotive Engineering Computer Science & IT Control & Systems Engineering Engineering Industrial & Manufacturing Engineering Materials Science & Nanotechnology

Description

International Journal on Orange Technologies (IJOT) is an online international peer-reviewed journal that publishes high-quality original scientific papers, short communications, correspondence, and case studies in areas of research, development, and applications of orange technology and ...