Platform for evaluating the next generation of DMS technology

p.u.l.s.e Overview

A platform to evaluate the next geneartion of DMS (Driver Monitoring System) technology for estimating driver status through eyelid motion and facial expresions obtained with facial recognition algorithms.

Driver status is estimated from camera, heartrate, voice, and vehicle movement data.

The accurate detection of driver status will be an exceedingly important aspect of the transition to Level 3 autonomous vehicles where it will be necessary to shift driving responsibilities from the car to the driver.

Target Market
Motor vehicles, ADAS, factory and office automation
Target Devices
Linux, Android, SBC (as exemplified by Raspberry Pi)

p.u.l.s.e Features

p.u.l.s.e uses data from both the driver and vehicle to estimate status. We're creating a platform that allows the user to collect data for evaluating using a driving simulator rather than an actual vehicle to ensure safety.

p.u.l.s.e Functionality

Relevant driving support is able to be provided by identifying the relationship between vehicle motion and the driver sleepiness and stress estimated using pulse and blink data.

p.u.l.s.e System architecture

p.u.l.s.e IR Sleepiness is estimated using blink frequency, etc., extrapolated from eye feature data through facial image recognition
p.u.l.s.e HR Sleepiness is also estimated using charactarist heart rate data gathered through a heartrate sensor
p.u.l.s.e VR Stress and illness are estimated by analyzing qualities in the driver's voice
p.u.l.s.e Sense

Driver status is estimated using the sensing data extrapolated from charactaristics in the vehicle and sensing data and indicators from various modules