Tackling The Menace of Cell Phone Distraction
In the first video, the driver is continually looking down. In the second video, the driver is looking at something on the phone while driving. In the third video, the driver is glancing down repeatedly. In the fourth video, the driver is talking on the phone.
While all 4 videos are examples of distracted driving, the way each of the 4 behaviors is detected is different from an AI perspective. Businesses and consumers alike understand the dangers of distracted driving and are keen to adopt technology that helps people stay attentive when they operate vehicles or machinery. There are many solutions in the market to detect risky driving and alert the driver, but they are all not created equally.
The most popular method of detecting distracted driving is using the head pose. Knowing the head pose and correlating it with the direction of the vehicle movement will help detect if the driver is not looking in the right direction – thus, distracted driving can be called out. This kind of distracted driving that is based on head pose is called owl distraction in DMS parlance – since the owl moves its head around.
If only head pose were used, distracted driving would not be detected in the other 3 cases!
In the second video, the head pose is aligned with the vehicle motion. However, there is a cellphone in the field of view which the AI can detect to alert the driver of distracted driving. This is how distracted driving is detected in the second and third cases. It is helpful to be able to differentiate between calling and texting since different drivers may be getting distracted with a different modality. Having a fine-grained visibility into distraction can make driver coaching more personalized and actionable.
If the AI were dependent on the head pose and cell phone for distracted driving, a distracted driver would not be called-out in such as the fourth case. The head pose does not change enough for it to be called distracted driving and the cellphone is not visible in the camera’s field of view.
In contrast to the owl distraction, there is a lizard eye distraction – called so because lizards do not move their heads to see elsewhere, they are able to do so just with glances. In the fourth video, we see an example of lizard eye distraction – frequent glances downward, but no appreciable head pose movement. Detecting lizard eye distraction is very hard – however, if one has to detect distracted driving comprehensively, this is a must.
LightMetrics DMS has extremely sophisticated AI that can detect distracted driving using many cues. No matter what the scenario is, if the driver is distracted, the driver will be alerted to pay attention to the road ahead. As a fleet manager, you can be assured that the technology is keeping an eye out for you on risky driving and will help you create a distracted driving free fleet.
To find out more about distracted driving and video telematics, please reach out to email@example.com. Follow us on LinkedIn to stay up to date with all the cool things happening at LightMetrics!