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 method, comprising:
receiving one or more images of a damaged vehicle, the damage vehicle comprising a plurality of normalized parts;
determining, using image segmentation on the one or more images, a location of damage on the damaged vehicle relative to one or more of the normalized parts; and
determining, using one or more classifiers, one or more damage states for one or more auxiliary parts of the vehicle based on at least the image segmentation.
2. The method of claim 1 , wherein the one or more classifiers comprises one or more of a first type of classifier and a second type of classifier,
wherein the first type of classifier is configured to perform image segmentation and determine a classification of damage to one of the normalized parts of the vehicle, and
wherein the second type of classifier is configured to i) receive input comprising image segmentation data and the one or more images and ii) generate output comprising the one or more damages states for the one or more auxiliary parts of the vehicle.
3. The method of claim 2 , wherein the segmentation data comprises one or more segmentation masks or segmentation maps.
4. The method of claim 1 , wherein the one or more classifiers perform the image segmentation on the one or more images.
5. The method of claim 1 , further comprising:
generating a plurality of cropped images from each of the one or more images, wherein determining the location of the damage on the vehicle further comprises using image segmentation on the plurality of cropped images.
6. The method of claim 1 , wherein the one or more classifiers comprises at least a multi-task classifier comprising both shared layers and task specific layers.
7. The method of claim 6 , wherein each task of the multi-task classifier correspond to a same auxiliary part.
8. The method of claim 1 , wherein determining, using the one or more classifiers, the one or more damage states for the one or more auxiliary parts of the vehicle is further based on a classification of a category of the damage.
9. One or more processors configured to perform operations comprising:
receiving one or more images of a damaged vehicle, the damage vehicle comprising a plurality of normalized parts;
determining, using image segmentation on the one or more images, a location of damage on the damaged vehicle relative to one or more of the normalized parts; and
determining, using one or more classifiers, one or more damage states for one or more auxiliary parts of the vehicle based on at least the image segmentation.
10. The one or more processors of claim 9 , wherein the one or more classifiers comprises one or more of a first type of classifier and a second type of classifier,
wherein the first type of classifier is configured to perform image segmentation and determine a classification of damage to one of the normalized parts of the vehicle, and
wherein the second type of classifier is configured to i) receive input comprising image segmentation data and the one or more images and ii) generate output comprising the one or more damages states for the one or more auxiliary parts of the vehicle.
11. The one or more processors of claim 10 , wherein the segmentation data comprises one or more segmentation masks or segmentation maps.
12. The one or more processors of claim 9 , wherein the one or more classifiers perform the image segmentation on the one or more images.
13. The one or more processors of claim 9 , further comprising:
generating a plurality of cropped images from each of the one or more images, wherein determining the location of the damage on the vehicle further comprises using image segmentation on the plurality of cropped images.
14. The one or more processors of claim 9 , wherein the one or more classifiers comprises at least a multi-task classifier comprising both shared layers and task specific layers.
15. The one or more processors of claim 14 , wherein each task of the multi-task classifier correspond to a same auxiliary part.
16. The one or more processors of claim 9 , wherein determining, using the one or more classifiers, the one or more damage states for the one or more auxiliary parts of the vehicle is further based on a classification of a category of the damage.
17. A non-transitory computer-readable storage medium storing a set of instructions that is executable by one or more processors, the set of instructions, when executed by the one or more processors, causing the one or more processors to perform operations, comprising:
receiving one or more images of a damaged vehicle, the damage vehicle comprising a plurality of normalized parts;
determining, using image segmentation on the one or more images, a location of damage on the damaged vehicle relative to one or more of the normalized parts; and
determining, using one or more classifiers, one or more damage states for one or more auxiliary parts of the vehicle based on at least the image segmentation.
18. The non-transitory computer-readable storage medium of claim 17 , wherein the one or more classifiers comprises one or more of a first type of classifier and a second type of classifier,
wherein the first type of classifier is configured to perform image segmentation and determine a classification of damage to one of the normalized parts of the vehicle, and
wherein the second type of classifier is configured to i) receive input comprising image segmentation data and the one or more images and ii) generate output comprising the one or more damages states for the one or more auxiliary parts of the vehicle.
19. The non-transitory computer-readable storage medium of claim 18 , wherein the segmentation data comprises one or more segmentation masks or segmentation maps.
20. The non-transitory computer-readable storage medium of claim 17 , wherein the one or more classifiers perform the image segmentation on the one or more images.
21. The non-transitory computer-readable storage medium of claim 17 , further comprising:
generating a plurality of cropped images from each of the one or more images, wherein determining the location of the damage on the vehicle further comprises using image segmentation on the plurality of cropped images.
22. The non-transitory computer-readable storage medium of claim 17 , wherein the one or more classifiers comprises at least a multi-task classifier comprising both shared layers and task specific layers.
23. The non-transitory computer-readable storage medium of claim 22 , wherein each task of the multi-task classifier correspond to a same auxiliary part.
24. The non-transitory computer-readable storage medium of claim 17 , wherein determining, using the one or more classifiers, the one or more damage states for the one or more auxiliary parts of the vehicle is further based on a classification of a category of the damage.