Image annotation information processing method, device, server and system

Abstract

The embodiments of this specification disclose a method, device, server and system for processing image annotation information, which can provide monitoring and judgment processing logic for different processing results of multiple nodes. Re-examination, revision, etc. In this way, the business ability of operators can be improved in the continuous review and feedback interaction, the efficiency of image annotation can be gradually improved, and the accuracy of image annotation in the training set can be greatly improved. The implementation of this description can effectively ensure the quality of labeling, and provide timely and effective information feedback in the workflow, thereby improving the efficiency of labeling information on sample images.

Description

technical field [0001] The embodiments of this specification belong to the technical field of computer data processing, and in particular, relate to a method, device, server and system for processing annotation information in images. Background technique [0002] The methods of detecting objects in images in existing applications mainly rely on the target detection / recognition models obtained through training, such as the parts recognition model in the vehicle damage assessment business, and the model for determining purchased items by taking pictures, etc. These target detection / recognition models usually need to rely on a large number of marked sample images for training, so the accuracy of the label information in the sample images has a great impact on the model output. [0003] When processing the annotation information of the sample image in the early stage, the sample image often contains multiple target subjects. The multi-target labeling task is more complicated th...

AI Technical Summary

Problems solved by technology

However, when the number of sample images to be processed is large and the target subjects in most images are relatively dense (at least two), manual labeling processing requires both professionalism and It is also easy to confuse, which often leads to low accuracy of the output sample image annotation information, and it is difficult to guarantee the quality of annotation, which in turn leads to poor image recognition algorithms and low target prediction accuracy in the image.

Images