Hackzurich project for T-Systems & Zurich Insurance
Smart devices have greatly reshaped our way of living, including mobility. Although they are bringing much convenience and benefits, the associated risks are also increasing: people are driving, walking, cycling, skiing, riding, etc. while being distracted by their smart devices.
Numerous facts support this statement, such as:
Fact 1: 43% of young people have walked into something or someone while checking their phone. In fact, countries are creating “smartphone” walking lanes goo.gl/N0Xt2H
Fact 2: 6 out of 10 car crashes involve distraction, out of which 12% are due to the use of mobile phones goo.gl/jVmPpT
This project intends to promote safe behaviours and new ideas to embed road-smartness into the future way of mobility.
- Microsoft Cognitive Services is used. MVP uses entity/item tagging to detect possible elements of the photo which may be assosiated with distraction / tiredness. i.e cellphone.
Custom Computer Vision Microsoft Azure offers the Custom Computer Vision service, which can be train with a set of photos of drives which may be sleepy / tired.
Similar face search We collect photos of people looking at the phone and use the Similar Face Search service to detect it.
Eye-Nose Distance If the head is directed towards down, it might be an indication for a driver sleeping or looking down at the phone, even if the phone may not be visible at a time. The distance between eyes and the nose, when projected onto a 2D-plane, is shorter in this case.
*/client Ionic app
*/server NodeJS Server
*/live-demo Live demo for the presentation
*/presentations Presentations, pitch decks etc.
- Deploy the server to T-Systems Cloud
- Integrate API in the Ionic App
- Integrate API in the Live-presentation
- Work on design
Adrian Barwicki ([email protected])
Ingmar Wolff (@TODO)
Marcel Engelmann ([email protected])
Lukas Maxeiner (@TODO)
MIT
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent