Method and apparatus for vehicle damage identification

An apparatus for vehicle damage identification, comprising:

• a processor;
• and
• a memory configured to store instructions, wherein the processor is configured to execute the instructions to:
• obtain a vehicle damage picture to be identified;
• input the vehicle damage picture into a plurality of pre-trained target detection models respectively, and obtain corresponding detection results from the plurality of pre-trained target detection models as a detection result set, wherein the plurality of pre-trained target detection models are configured to perform detection on the vehicle damage picture in parallel, and the detection result set comprises candidate bounding boxes detected by the plurality of pre-trained target detection models;
• determine an integrated feature vector of a first candidate bounding box of the candidate bounding boxes according to an association relationship between the first candidate bounding box and other candidate bounding boxes, wherein the association relationship includes at least one of:
• a bounding box content association relationship, a location association relationship, or a category prediction result association relationship;
• and
• separately input integrated feature vectors corresponding to the candidate bounding boxes into a pre-trained classification model, and optimize the detection result set according to output results of the pre-trained classification model.

A non-transitory computer-readable medium having stored therein instructions that, when executed by a processor of a device, cause the device to perform a method for vehicle damage identification, the method comprising:

• obtaining a vehicle damage picture to be identified;
• inputting the vehicle damage picture into a plurality of pre-trained target detection models respectively, and obtaining corresponding detection results from the plurality of pre-trained target detection models as a detection result set, wherein the plurality of pre-trained target detection models are configured to perform detection on the vehicle damage picture in parallel, and the detection result set comprises candidate bounding boxes detected by the plurality of pre-trained target detection models;
• determining an integrated feature vector of a first candidate bounding box of the candidate bounding boxes based on an association relationship between the first candidate bounding box and other candidate bounding boxes, wherein the association relationship includes at least one of:
• a bounding box content association relationship, a location association relationship, or a category prediction result association relationship;
• and
• separately inputting integrated feature vectors corresponding to the candidate bounding boxes into a pre-trained classification model, and optimizing the detection result set according to output results of the pre-trained classification model.