Car Damage Severity Assessment Using Supervised Deep Learning

DOI

10.2991/978-94-6463-589-8_28How to use a DOI?

Keywords

supervised deep learning; image recognition; transfer learning; car damage; EfficientNet

Abstract

The assessment of car damage severity is a critical task in the automotive industry, particularly in insurance claims processing, repairs, and traffic safety analysis. Traditionally, car damage assessment has heavily relied on manual inspections by experts. However, this approach suffers from several limitations, including subjectivity, time consumption, and potential errors, which can lead to inaccurate assessments, delayed insurance claims, and compromised repair processes. This study aims to addresses challenges in accurately assessing car damage severity using deep learning techniques, aiming to overcome limitations associated with manual inspection, such as subjectivity, time consumption, and potential errors. A robust data preprocessing pipeline is implemented using TensorFlow's ImageDataGenerator to prepare and augment the car damage severity assessment dataset, enhancing the model's ability to generalize across diverse data. Once preprocessed, the study continues with 4 supervised deep learning model which are Roboflow, ResNet, EfficientNetV2L and VGG19, with the best performing is EfficientNetV2L with an accuracy of 81%. When trained on 400x400 pixel images for 10 epochs.

Copyright

© 2024 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Ahmad Taufiq Mohamad
AU  - Nur Atiqah Sia Abdullah
AU  - Suzana Ahmad
AU  - Garry Baldwin Francis
PY  - 2024
DA  - 2024/12/01
TI  - Car Damage Severity Assessment Using Supervised Deep Learning
BT  - Proceedings of the International Conference on Innovation & Entrepreneurship in Computing, Engineering & Science Education (InvENT 2024)
PB  - Atlantis Press
SP  - 308
EP  - 318
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6463-589-8_28
DO  - 10.2991/978-94-6463-589-8_28
ID  - Mohamad2024
ER  -