The project boost Intelligent Functions
to Improve in -vehicle UX, AVA: Automotive Virtual Assistant.
In the particular case of camera-based smart sensing system for vehicle compartment, the final goal of AVA is to implement a low-cost solution based on a VPU package on an embedded syste, that support already pre-trained DNN with recognition capabilities as check seat belt, mobile distractions, view on the road, eyes closed [prevent fatigue] to increase safety on-board. To be integrated either in the rearview mirror or in the sunvisors, even in the cockpit.
OB1.: to define an AI vision architecture and the related basic buiding blocks to implement a highly scalable system compliant with the system requirements.
OBJ2.: to take advantatge of the state-of-the-art AI embedded cameras and image/video processing technologies for deep learning.
OBJ3.:increase safety and comfort on-board, transparent to the final user
OBJ4.:to obtain an affordable AI platform suitable for use in all kinds of image recognition vehicle compartment applications.