Alter is a video based embedded solution for detecting drivers’ signs of fatigue and tendency to fall asleep during driving in real-time using AI technology that assesses the driver’s drowsiness by monitoring a clinical sleep indicator known as Karolinska Sleepiness Scale or KSS. The KSS indicator includes a set of parameters derived from facial expressions captured through by the AI model.
* Accuracy exceeds 96.5%
* Trained on over 50,000 hours of video datasets.
* Classifies the drivers state of sleepiness into 5 states: alert, semi-alert, moderately-sleepy, semi-sleepy, sleepy.
* Immune to poor lighting conditions.
* Works with various head poses and orientations.
* Triggers an audio alert as well as an SMS message configurable for various levels of sleep.
* Operate in real-time or on recorded videos.
* Compatible with most digital cameras.