Auxiliary parts damage determination
A method of determining, one or more damage states of one or more auxiliary parts of a damaged vehicle, the vehicle comprising a plurality of normalized parts and at least some of the normalized parts further comprising one or more auxiliary parts. The method includes receiving one or more images of the vehicle, using a plurality of classifiers, each determining at least one classification of damage to the vehicle, each said classification being determined for each of a plurality of normalized parts of the vehicle, determining one or more classifications for the plurality of auxiliary parts using one or more trained models, wherein each classification comprises at least one indication of damage to at least one auxiliary part and outputting the determined damage states of the one or more auxiliary parts.
1. A computer-implemented method of determining one or more damage states of one or more auxiliary parts of a damaged vehicle, the vehicle comprising a plurality of normalized parts and at least some of the normalized parts further comprising one or more auxiliary parts, comprising:
receiving one or more images of the vehicle;
using a plurality of classifiers, each determining at least one classification of damage to the vehicle, each said classification being determined for each of a plurality of normalized parts of the vehicle;
determining one or more classifications for the plurality of auxiliary parts using one or more trained models, wherein each classification comprises at least one indication of damage to at least one auxiliary part; and
outputting the determined damage states of the one or more auxiliary parts.
2. The computer-implemented method of claim 1 , further comprising determining one or more auxiliary parts corresponding to each of the plurality of normalized parts of the damaged vehicle.
3. The computer-implemented method of claim 1 , wherein the one or more trained models comprise any one or any combination of: a neural network; a convolutional neural network; a recurrent neural network; and/or multi-task learning network.
4. The computer-implemented method of claim 1 , wherein the one or more trained models are trained on each of the normalized parts of a vehicle.
5. The computer-implemented method of claim 1 , further comprising determining a plurality of normalized parts of the vehicle that are represented in the plurality of images of the vehicle comprising the use of a plurality of classifiers.
6. The computer-implemented method of claim 5 , wherein each one of the plurality of classifiers is operable to detect each of the normalized parts of the vehicle, optionally wherein each one of the plurality of classifiers is further operable to determine a damage classification.
7. The computer-implemented method of claim 1 , wherein said plurality of normalized parts of the vehicle comprise any or any combination of: normalized regions of the vehicle; normalized zones of the vehicle; standardized parts of the vehicle;
standardized regions of the vehicle; standardized zones of the vehicle.
8. The computer-implemented method of claim 1 , wherein the one or more auxiliary parts comprise any or any combination of: peripheral components of the normalized parts; subparts of the normalized parts; and/or non-standardized vehicle parts specific to any or any combination of a predetermined make, model and year of vehicle.
9. The computer-implemented method of claim 1 , wherein determining one or more classifications for the plurality of auxiliary parts further comprises an additional trained classifier to compare each classification for the determined each of a plurality of normalized parts of the vehicle and each of the received one or more images of the vehicle.
10. The computer-implemented method of claim 9 , wherein the additional trained classifier is trained on any or any combination of: multiple visible auxiliary parts; multiple hidden auxiliary parts; standardized auxiliary parts; manufacture specific auxiliary parts; jurisdictional specific auxiliary parts.
11. The computer-implemented method of claim 1 , wherein the one or more trained models or one or more classifiers arc arranged as task specific layers and/or shared layers.
12. The computer-implemented method of claim 1 , further comprising correlating the damage classification of each of the plurality of normalized parts to corresponding auxiliary parts for the or each of the normalized part.
13. The computer-implemented method of claim 1 , wherein determining one or more classifications for the plurality of auxiliary parts comprises the use of the damage classification of each of the plurality of normalized parts as input for the one or more trained models.
14. The computer-implemented method of claim 1 , wherein determining at least one classification of damage to the vehicle further comprises generating one or more segmentation maps of the damage.