Image recognition method and device and computer readable storage medium

Abstract

The embodiment of the invention discloses an image recognition method and device and a computer readable storage medium. According to the embodiment of the invention, the method comprises the steps of after an image sample set is acquired, adopting a preset recognition model to perform feature extraction on the image samples in the image sample set to obtain an image feature set, then using the image samples as data nodes to construct the neighbor graph according to the image feature set, and then correcting the basic labels of the image samples based on the neighbor graph; obtaining a corrected image sample set, then training a preset recognition model by adopting the corrected image sample set, and recognizing the to-be-recognized image through the trained recognition model. According to the scheme, the image recognition accuracy can be improved.

Description

Technical field [0001] The present invention relates to the field of communication technologies, and more particularly to an image identification method, a device, and a computer readable storage medium. Background technique [0002] In recent years, with the high heat of neural network technology in the artificial intelligence, it has also been made in terms of image recognition, especially the pedestrians in the image. During the reconciliation process of pedestrian, it is necessary to label the image samples. Since the label is difficult, it is often caused to have noise in the training data set, so noise is required. The existing image identification method tends to use a classifier to remove noise on the predictive value of the training sample, and in turn, the neural network is trained according to the denoising training data set to identify the image. [0003] In the process of research and practice in the prior art, the inventors of the present invention find that the cla...

AI Technical Summary

Problems solved by technology

[0003] In the process of research and practice of the prior art, the inventors of the present invention found that when the number of image samples of each identified pedestrian is small, the classifier is very sensitive to labeling errors, so that the classifier is very sensitive to the image The accuracy and recall of samples for denoising are very low, therefore, the accuracy of image recognition is greatly reduced

Images