Training method and system of neural network model with anti-counterfeiting function, anti-counterfeiting verification method and electronic device

Abstract

The invention relates to a training method and system of a neural network model with an anti-counterfeiting function, an anti-counterfeiting verification method and an electronic device. In the training method and system provided by the invention, the data with the preset sampling ratio in the training data set is replaced with the data with the preset identifier, the initial feature of the data with the preset identifier is changed into the preset error feature, and finally, the training data set with the preset error feature is used for training the model. According to the method and the system, the neural network model can have an anti-counterfeiting function and an anti-theft function on the premise of not reducing the performance of the model. In the anti-counterfeiting verification method, after model training is completed, data without a specific identifier can be normally operated and a result is output, and for a trigger sample with a preset identifier, a preset error result occurs, so that anti-counterfeiting verification is realized. The electronic device also has the same beneficial effects as the training method of the neural network model with the anti-counterfeitingfunction.

Description

【Technical field】 [0001] The invention relates to the field of artificial intelligence, in particular to a training method and system for a neural network model with anti-counterfeiting functions, an anti-counterfeiting verification method and an electronic device. 【Background technique】 [0002] With the rapid development of artificial intelligence, deep neural networks have achieved great success in the fields of image recognition, speech recognition and natural language processing. Many institutions build artificial intelligence systems or artificial intelligence services based on deep neural network models. Since the training of the neural network model requires the collection and labeling of a large amount of data and the use of a large amount of computing resources, in addition, the design of the model also requires a large amount of machine learning expertise and a large number of repeated experiments, which also requires a lot of manpower, material resources and time...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem that the existing neural technology model is stolen and it is difficult to verify its ownership, the present invention provides a training method of a neural network model with anti-counterfeiting function and its system, an anti-counterfeiting verification method and an electronic device

Images