Specific scene model upgrading method and system based on federated learning

Abstract

The invention discloses a specific scene model upgrading method based on federated learning. The method comprises the following steps: selecting an AI terminal; marking all scene data in a marking module of each AI terminal, storing the scene data in the local of the terminal, calling an intelligent classification and statistics module to perform classification statistics on the scene data, and reporting the scene data to an alliance statistics module; forming a federated learning alliance by the terminals similar to the usage scenarios, wherein after a preset updating condition is met, the alliance statistical management module triggers and starts transverse federation learning training of the updating model; screening the training data before the training of the updating model; performing a learning updating process by the transverse federation; in the process, the preset updating condition can be modified along with the increase of the number of terminals and the increase of data volume, and iterative updating is continued. According to the invention, the problems of low model updating and upgrading efficiency of the artificial intelligence terminal and low accuracy in an actualuse scene can be solved.

Description

technical field [0001] The invention relates to the field of federated learning for artificial intelligence recognition, in particular to a method and system for upgrading a specific scene model based on federated learning. Background technique [0002] At present, most AI models are trained on the server with marked samples, and after passing the test set test, they are transplanted to the terminal device. Because data collection and labeling are cumbersome and involve privacy and security, AI models are often updated after a certain amount of data is collected next time or the accuracy of the model cannot adapt after deployment. The new model repeats the above deployment process, that is, retrains on the server with newly collected and labeled data, and then replaces the original model on the terminal device to complete the update step. [0003] There are several key points in the above process: [0004] (1) Data acquisition: The current AI algorithm model is very depend...

AI Technical Summary

Problems solved by technology

[0031] The purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art, and provide a specific scene model upgrade method based on federated learning, which uses the specific scene data and labeling information collected by the artificial intelligence terminal alliance similar to the application scene, combined with horizontal federation Learning technology, retraining and updating the upgraded model; this method solves the problems of low efficiency of model updating and upgrading of artificial intelligence terminals and low accuracy in actual use scenarios

Images