A system automatic review method based on multi-person collaborative image annotation

Abstract

The invention discloses a system automatic check method based on multi-person cooperative image annotation. The method can automatically check the overall image annotation quality, avoid the trouble of special annotation by professionals and greatly reduce the check workload. The method avoids the check step by professionals, and for developers, automatic checking improves the work efficiency andreduces the workload. In addition, evaluation of the data quality is stricter, which lies in that the algorithm of the invention evaluates all annotation data. Moreover, annotation errors are greatlyreduced and thus the annotation quality is improved through multi-person joint annotation.

Description

technical field [0001] The invention belongs to the technical field of data labeling, and in particular relates to a system automatic review method based on multi-person collaborative image labeling. Background technique [0002] Machine learning is the study of how computers simulate and realize human learning behavior. It is the core of artificial intelligence and has a wide range of applications and immeasurable prospects. The most widely used machine learning is supervised learning, but there is a fatal flaw in supervised learning, which requires a large amount of manual labeling data to make the learning effect better. However, labeling useful data in massive data requires huge The workload is extremely high and unrealistic to be marked by developers and researchers (relatively few personnel), so a large number of professional labeling companies and crowdsourcing platforms have emerged to provide corresponding services. However, a serious problem has arisen from this. ...

AI Technical Summary

Problems solved by technology

[0005] For the image annotation tasks of object detection and object region segmentation, one of the current audit solutions is to add labeled standard sample data when distributing annotation tasks to annotators to evaluate the annotators and measure their annotation quality, but Labeled standard data cannot be easily obtained, and special labeling by professionals is required, which undoubtedly increases the audit cost; moreover, since only a small part of the data is specially labeled, the overall quality can only be evaluated by part of the labeling quality

Images