Data-augmented deep semi-supervised extreme learning image classification method and system

Abstract

The invention discloses a data-augmented deep semi-supervised extreme learning image classification method and system. The method comprises: performing feature extraction on a training image by adopting a deep convolutional network model; finely adjusting and optimizing the deep convolutional network model based on part of artificial label data, and generating a pseudo label for the training image without the label; fusing the high-level semantic features extracted from the training image with the low-layer and shallow-layer structural features to obtain fused image features; augmenting the fused image features and labels of the training images by adopting a random linear interpolation technology; and training a single hidden layer feedforward neural network for the augmented fusion image features and labels, and replacing a full connection layer in the deep convolutional network model to obtain a final image classification and recognition network model. The invention has the advantages of low manual marking requirement, robust noise interference resistance, good classification and recognition performance, high task expansibility and data augmentation.

Description

technical field [0001] The invention relates to the technical fields of image classification and target recognition, in particular to a data-augmented deep semi-supervised over-limit learning image classification method and system. Background technique [0002] Most of the current visual target recognition methods with better performance use deep learning technology, which often requires supervised learning and training of deep neural network models on large-scale manually labeled data. However, manually labeling large amounts of data is expensive and even impractical for some application scenarios. In recent years, related research has focused more on deep semi-supervised learning techniques. Deep semi-supervised learning techniques can leverage a small amount of high-quality labeled data and a large amount of easily available unlabeled data to improve the performance of visual object classification and recognition methods. At present, a variety of deep semi-supervised le...

AI Technical Summary

Problems solved by technology

In addition, this technology is also difficult to effectively set the classifier, the utilization efficiency of sample data information is low, and it is difficult to carry out efficient learning and training

Images