Extreme Inception Deep Learning Model for Multi-Factor Car Dent Severity Assessment and Repair Cost Estimation

Traditionally, the evaluation of damage to the vehicle structure is conducted subjectively, which might yield conflicting results and diminish consumer trust. This study addresses the problem by employing a cutting-edge deep learning methodology utilizing the Xception model for damage assessment. This method employs the pre-trained and fine-tuned Xception model alongside feature engineering. The model can categorize damage as minimal, moderate, or severe by analyzing photographs of the damage. The model demonstrated exceptional capability to provide precise outcomes despite a limited dataset of 1,436 samples, attaining a validation accuracy of 83%. Additionally, we have created an intuitive web API that accepts an image of the damage, along with the model, make, year, and mileage of the vehicle, to provide an evaluation of the car's value based on the extent of the damage. This approach ensures precise, efficient, and transparent outcomes for evaluating the expenses of automotive repairs and analyzing damage.