An automatic image annotation method for weakly supervised semantic segmentation

Abstract

An automatic image annotation method for weakly supervised semantic segmentation. The object border is located by an image object detection method, and the semantic label is given. The object border and the semantic label are regarded as a kind of weak supervised semantic label of image level. By using traditional image segmentation method, the whole object region is segmented out, and the segmentation template for training classification network is generated. Then, the segmentation template is used as a supervisory signal to train the classification network. Finally, the trained classification network is used to segment the test image semantically. The technical proposal of the invention utilizes an object detection method to obtain a border and a semantic tag of an object in an image, utilizes a traditional image segmentation method to segment an object region, and combines the semantic tag to serve as a training sample for weak supervision semantic segmentation. The method to automatically generate training samples for weak supervised semantic segmentation, solves the problem of time-consuming and laborious manual labeling of a large number of images.

Description

technical field [0001] The present invention relates to the technical field of semantic segmentation based on deep learning, and more specifically, to an automatic image labeling method for weakly supervised semantic segmentation. Background technique [0002] Traditional image segmentation algorithms use an interactive method to segment the target area by specifying the foreground and background areas or pixels by the user. Generally, there are two types of user-specified interaction methods, one is a trimap with a large amount of information, and the other is a sketch (scribbles) with a small amount of information. The tripartite map is a graph that contains the foreground, background and unknown regions made manually. The structure of the tripartite map is complex, and it is almost necessary to fill the entire image manually. The sketch is much more convenient, as long as a few lines are drawn on the foreground and background. Just a pen. However, all these image annota...

AI Technical Summary

Problems solved by technology

Manual data labeling is a time-consuming and laborious work. At present, there is no unified tool for image labeling, and it is easily affected by the personal factors of the observer. It is generally difficult to achieve an accuracy of more than 90% for manually labeled training samples (that is, the labeling is correct. Rate)

If the accuracy of manual labeling is too low, it will directly lead to machine prediction errors

Images