IVAA: Intelligent Vehicle Accident Analysis System
Kundjanasith Thonglek, Norawit Urailertprasert, Patchara Pattiyathanee, Chantana Chantrapornchai · 2019
Intelligent Vehicle Accident Analysis (IVAA) system provides an artificial intelligence as a service (AIaaS) for building a system that can automatically evaluate vehicle parts' damage from the accident. Particularly, IVAA can recognize the damaged vehicle part and the severity level facilitating the insurance company's claiming process. There are four main elements in the system which support four stakeholders in an insurance company: insurance experts, data scientists, operators and field employees. Insurance experts utilize the data labeling tool to label damaged parts of a vehicle in a given image as a training data building process. Data scientists iterate to the deep learning model building process for continuous model update. Operators monitor the visualization system for daily statistics related to the number of accidents based on locations. Field employees use LINE Official integration to take a photo the damaged on the vehicle at the accident site and retrieve the repair estimation. IVAA is built on the docker image which can scale-in or scale-out the system depend on utilization efficiently. We deploy the Faster Region-based convolutional neural network, along with residual Inception network to localize the damage region and classify into 5 damage levels for a vehicle. The accuracy of the localization is 93.28 percents and the accuracy of the classification is 98.47 percents.