Skip to Content

Whitepaper: Assessing Vehicle Damage – the ML way

Sogeti Labs
May 01, 2019

In this research, we propose to replace the traditional methods of vehicle damage inspection (for car insurance) with a Machine Learning based automated solution.

Deep Learning based models are used to recognize whether a car in a given image is damaged and its type/severity. A promising attempt in classifying car damages into multiple classes is presented. The emphasis is on fine-tuning this model iteratively with the objective of achieving satisfactory accuracy scores.

This research open doors for future collaborations on image recognition projects in general and for the car insurance field in particular by enabling insurers to assess customer vehicle damage and expedite claims settlement.

Download the report here.

About the author

SogetiLabs gathers distinguished technology leaders from around the Sogeti world. It is an initiative explaining not how IT works, but what IT means for business.


    One thought on “Whitepaper: Assessing Vehicle Damage – the ML way

    1. Interesting. For Liberty Insurance in Ireland we have researched exactly the same technology in a Thinkubator workshop and presented some possibilities in January last year. Have you been in touch with John Mcintyre as well?

    Leave a Reply

    Your email address will not be published. Required fields are marked *

    Slide to submit