METHOD OF DETERMINING REPAIR OPERATIONS FOR A DAMAGED VEHICLE INCLUDING USING DOMAIN CONFUSION LOSS TECHNIQUES

1 - 18 . (canceled)

19 . A computer-implemented method for estimating damage to a vehicle, comprising:

receiving a plurality of images of the vehicle;

determining a plurality of parts of the vehicle that are represented in each of the plurality of images of the vehicle;

determining one or more sets of relevant images of each of the plurality of parts from the plurality of images of the vehicle;

determining using one or more trained models a damage state of each of the parts of the vehicle using the one or more relevant images of each of the plurality of parts of the vehicle, the one or more damage states being determined as one or more quantitative values, wherein the one or more trained models comprise a domain confusion loss; and

outputting the one or more damage states of each of the parts of the vehicle.

20 . The computer-implemented method of claim 19 , further comprising:

receiving vehicle input data, wherein the vehicle input data comprises details of one or more proposed parts and labor operations for repairing the damage to the vehicle; and

determining the parts and labor operations of the vehicle input data that are relevant to each of the plurality of normalized parts of the vehicle.

21 . The computer-implemented method of claim 19 , further comprising determining one or more scores indicative of a damage value based one or more classifications indicating damage to at least one part performed by the one or more trained models.

22 . The computer-implemented method of claim 21 , wherein determining one or more repair operations for each of the parts of the vehicle comprises comparing the one or more scores to a pre-determined threshold.

23 . The computer-implemented method of claim 19 , further comprising the use of any or any combination of: a computer vision damage assessment model; a repair/replace prediction model; a labor hours prediction model; and/or a remove and install model.

24 . The computer-implemented method of claim 23 , wherein the any or any combination of: a computer vision damage assessment model; a repair/replace prediction model; a labor hours prediction model; and/or a remove and install model are used as one or more secondary models that receive input from said one or more trained models.

25 . The computer-implemented method of claim 19 , wherein the one or more repair operations for each of the parts of the vehicle comprise replacing or repairing the damaged part of the vehicle.

26 . The computer-implemented method of claim 19 , further comprising determining repair or replace labor hours based on the one or more classifications.

27 . The computer-implemented method of claim 19 , wherein the pre-determined threshold is adjustable based on one or more jurisdictional requirements.

28 . The computer-implemented method of claim 19 , further comprising determining a plurality of parts of the damaged vehicle that are represented in the plurality of images of the damaged vehicle comprise the use of a plurality of classifiers.

29 . The computer-implemented method of claim 28 , wherein each one of the plurality of classifiers is operable to detect each of the parts of the damaged vehicle.

30 . The computer-implemented method of claim 29 , further comprising determining one or more sets of relevant images of each of the plurality of parts from the plurality of images of the vehicle.

31 . The computer-implemented method of claim 19 , wherein the one or more trained models comprise one or more of: a neural network; a convolutional neural network; and/or a recurrent neural network.

32 . The computer-implemented method of claim 19 , wherein the one of more trained models comprises one or more networks trained on a plurality of datasets and wherein the one or more trained models are operable to determine a proximity between different data points in a feature space, optionally wherein one or more similar data points from are selected.

33 . The computer-implemented method of claim 19 , further comprising querying one or more databases to determine any of: pre-painted vehicle part costs; additional cost for a pre-painted part; total paint time for a pre-painted part; paint labor cost; material cost; and/or total paint cost.

34 . A computer-implemented method of training a neural network for estimating damage to a vehicle, including neurons, each neuron being associated with weights, wherein the method comprises:

obtaining training inputs, the training inputs comprising images of the damaged vehicles and whether any parts of the vehicles are damaged;

for each training input, selecting one or more neurons based on their respective probability;

adjusting the weights of the selected neurons such that the selected neurons are substantially operable to classify, per part of the vehicle, whether each part is damaged;

processing the training input with the neural network to generate a predicted output; and

adjusting the weights based on the predicted output.

35 . The computer-implemented method of claim 34 , wherein the neural network comprises a visual model and/or computer vision model.

36 . The computer-implemented method of claim 34 , wherein the neural network is trained on data from a smaller dataset and a subset of data from a larger dataset, the subset of data from the larger dataset comprising the most similar data points to the smaller dataset.