Training method and device of image classification model

Abstract

The invention provides a training method and device of an image classification model. The method comprises the steps of: obtaining class center features of at least two classes corresponding to an image sample set comprising noise image samples and features of all the image samples in the image sample set, and marking the image samples with original class labels; determining the similarity betweenthe class center features of at least two classes and the features of each image sample; for each image sample, taking the category to which the category center feature corresponding to the maximum similarity belongs as a new category label of the corresponding image sample to perform sample labeling to obtain a target image sample labeled with an original category label and a new category label;constructing a loss function of an image classification model based on the original category label, the new category label and the determined similarity; and training an image classification model byadopting the target image sample based on the loss function. According to the invention, the prediction accuracy of the image classification model obtained by training can be improved.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence, in particular to a training method and device for an image classification model. Background technique [0002] Artificial intelligence technology is a comprehensive subject that involves a wide range of fields, including both hardware-level technology and software-level technology. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology, and machine learning / deep learning, etc. Several major directions. Among them, machine learning (ML, Machine Learning) is the core of artificial intelligence and the fundamental way to make computers intelligent, and its application pervades all fields of artificial intelligence. Machine learning and deep learning typically include techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, and induct...

AI Technical Summary

Problems solved by technology

Due to the mistakes of the labelers, unclear concepts, etc., the image sample set usually includes noise image samples, and the classification prediction performance of the image classification model trained based on the image data set including noise image samples is low.

Images