Image processing method and device and storage medium

Abstract

The embodiment of the invention provides an image processing method and device and a storage medium. The method comprises the steps of obtaining a first image which comprises at least one time domain feature and at least one frequency domain feature and is obtained according to first voice data to be processed, and updating the first image according to a preset strategy to obtain a second image, wherein the preset strategy at least comprises one of the following items: transforming a time domain position of at least one time domain feature on a time domain, or transforming the frequency domain position of at least one frequency domain feature on the frequency domain. According to the scheme, rich and diversified training data can be provided for the training model, so that the training effect is improved.

Description

technical field [0001] The embodiments of the present application relate to the technical field of data processing, and in particular, to an image processing method, device, and storage medium. Background technique [0002] Neural networks are widely used in tasks such as image recognition and text analysis. Generally, neural networks need to be trained based on large-scale data. In order to expand the training data, an additional neural network model (for example, using an autoencoder model) is generally trained, and the additional neural network model is used to synthesize new speech data, that is, the existing speech data is converted into new speech data with a similar structure, so that Expand the amount of voice data. [0003] During the research and practice of the prior art, the inventors of the embodiments of the present application found that although for image classification tasks, millions of pictures and documents can be downloaded, processed and labeled from t...

AI Technical Summary

Problems solved by technology

[0003] During the research and practice of the prior art, the inventors of the embodiments of the present application found that although for image classification tasks, millions of pictures and documents can be downloaded, processed and labeled from the Internet to construct data sets, but for As far as voice data is concerned, it takes a lot of resources in the collection process (such as manual recording of voice data) and preprocessing process, and it is easy to label errors.

It can be seen that the existing expansion schemes for training data are generally efficient and the richness of expansion is limited.

Images