Personalized Autonomous Vehicle and ADAS
Autonomous Vehicle: Enhancing Trust and Comfort through AI-Driven Driver Behavior Prediction
In the evolving field of autonomous vehicles and Advanced Driver-Assistance Systems (ADAS), personalization is often overlooked. Driving is inherently a personalized experience; each of us has unique driving habits and preferences. For instance, some drivers prefer to slow down near curves, while others maintain their speed. Without personalizing autonomous vehicles and ADAS, it becomes challenging to establish the trust and comfort we feel when driving ourselves or being driven by someone familiar.
Our project, developed during the Ford Road Trip Hackathon, addresses this gap by personalizing autonomous vehicles and ADAS systems. We aimed to create an experience that closely mimics individual driving behaviors, thereby enhancing trust and comfort in these advanced technologies.
Project Overview:
Our AI-based system predicts driver behavior to personalize ADAS and autonomous vehicle responses. Utilizing Long Short-Term Memory (LSTM) networks, Convolutional Neural Networks (CNN), and an innovative visual cortex training system, our approach integrates human driving experience and knowledge to accurately predict driving behaviors.