Manually assessing hail damages on a car bodywork is a tedious and time-consuming process. For our client, a car damage repair company, we developed a computer vision based solution that automates this process with 90% accuracy, reducing the assessment time from 40 minutes to less than 5 minutes per car. In this talk, we will present our machine learning solution and describe how we handled challenges like data scarcity and processing time constraints.

Download the slides for this talk.Download ( PDF, 8074.47 MB)