Image processing method and device based on artificial intelligence, and storage medium

Abstract

The embodiment of the invention provides an image processing method and device based on artificial intelligence and a storage medium, and the method comprises the steps: a training sample set, which comprises a first image sample set with a class label and a second image sample set composed of triples, is acquired; semantic learning training and metric learning training are conducted on an original model through the training sample set, the original model comprises a first branch network and a second branch network, and the first branch network and the second branch network comprise shared network parameters; an image feature extraction model is determined according to the trained original model, and the image feature extraction model is used for extracting the feature vectors of the images, so that the model has the semantic extraction capability while realizing metric learning, and the accuracy of image retrieval can be improved based on the image features extracted by the model.

Description

Technical field [0001] The present invention relates to computer technology, and in particular relates to an image processing method based on Artificial Intelligence, apparatus and a storage medium. Background technique [0002] With the rapid development of Internet technology, the rapid growth in the amount of multimedia data, particularly image data as the main information bearer, how to accurately retrieve and query and retrieve images similar to the image data in these vast amounts of images, it has become a research hot spots. [0003] Image retrieval capabilities often depends on the image feature characterizing an image, the image in the mass data driving scene, the conventional image contrast for the image feature, such as a scale invariant feature transform (Scale-invariant featuretransform, SIFT) features, color histogram wherein, the gradient direction histogram (histogram of OrientedGradient, HOG) image of the features like poor performance when processing the image ...

AI Technical Summary

Problems solved by technology

[0003] The ability of image retrieval often depends on the image features that characterize the image. In the image-driven scenario of massive data, the traditional image features used for image comparison, such as scale-invariant feature transform (SIFT) feature, color histogram Image features such as features and histogram of oriented gradients (Histogram of Oriented Gradient, HOG) features do not perform well when dealing with unseen images or noisy images, etc. Image features extracted by deep learning models based on metric learning, such as SimCLR ( Simple Framework for Contrastive Learning of Visual Representations) method and BYOL (Bootstrap Your OwnLatent A New Approach to Self-Supervised Learning) method, in processing images that have undergone certain space transformation attacks (such as cropping more than 1 / 3) or have never seen image, often cannot accurately and effectively retrieve the image data

Images