Abstract: Image and video-based vehicle insurance claim processing is an important area with a large scope for automation in the insurance sector. At present, the insurance claim process and validation are done manually resulting in a time-consuming process with less scalability and more prone to error. Especially during the pandemic times, manual inspection is a difficult process and claim amounts primarily rely on the type of damage and damaged part of the car, so rise the need for an automated system for the whole process of car insurance claim as which can efficiently classify and detect damage and helps to minimize the claim leakage. Also, there is a chance of faking car damage images using image forgery or deepfake generation techniques. To analyze and design an automated vehicle insurance claim platform that can perform car damage detection and classification along with image forgery & deepfake detection.
Keywords: Transfer learning, YOLO, image copy-move forgery detection (CMFD), speeded-up robust feature (SURF), polar complex exponential transform (PCET), Deepfakes, LSTM
| DOI: 10.17148/IJARCCE.2021.10527