End-to-end weak supervision target detection method based on salient guidance

Abstract

The invention discloses an end-to-end weak supervision target detection method based on salient guidance. According to the method, a deep neural network is constructed, and a salient sub-network of a target box is added on the basis of a weak supervised classifier network; meanwhile, seed target areas with related categories are selected according to the criterion of context difference by the aid of a category-related salient map trained with a weak supervision method, and the salient sub-network and a classifier sub-network are supervised and trained. Compared with existing weak supervision target detection methods, the method has the advantages that better performance is obtained, meanwhile, only training of image-grade labels is required, and the workload of training data labeling is reduced.

Description

technical field [0001] The invention relates to an image target detection method, in particular to an end-to-end weakly supervised target detection method based on saliency guidance. Background technique [0002] Object detection aims to mark the objects and their categories in the image with rectangular boxes. Traditional machine learning-based object detection methods generally belong to the supervised learning method, which requires training data with the true value of the rectangular box. However, in big data scenarios, marking a large number of rectangular boxes requires a lot of manpower, which limits the application of supervised learning methods on large-scale data sets. [0003] In order to solve the problem of expensive training data, in recent years, researchers have developed learning methods based on weakly supervised labels. Although the supervision ability of weakly supervised marks is weaker than that of supervised marks, the cost of obtaining weakly superv...

AI Technical Summary

Problems solved by technology

However, these methods only use the features in the deep neural network to obtain more information in the image, so the performance of target detection is not greatly improved.

Images