City defense and control image detection method and system based on asynchronous federated learning

Abstract

The invention discloses a city defense and control image detection method based on asynchronous federated learning. The method comprises the steps that a cloud server initializes a global model; each end device carries out scene labeling on city defense and control image data, divides the city defense and control image data into a training set and a test set, and screens out a data set which is close to the sample distribution of the test set from the training set as a to-be-trained set; a global model is obtained from the cloud server, and the to-be-trained set is locally trained to obtain a local model; each end device uploads the local model to the cloud server after homomorphic encryption; and the cloud server performs global model training on the local model by adopting an asynchronous federation calculation strategy to obtain an updated global model.

Description

technical field [0001] The invention relates to the technical field of image recognition, in particular to an image detection method and system for urban prevention and control based on asynchronous federated learning. Background technique [0002] At present, image recognition scenarios for urban prevention and control mainly include application scenarios such as road construction damage, illegal parking of vehicles, floating objects on the river surface, and garbage dumping. Since the above application scenarios are widely distributed in various parts of the city, collaboration between different companies is required. For example, company A uses a camera for image data collection, and company B uses a drone to mount a camera for image data collection. Since the pictures collected by different companies cannot be shared directly due to reasons such as data security and privacy protection, it is necessary to establish a federated training network. Each company extracts data...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a method and system for urban prevention and control image detection based on asynchronous federated learning, which aims to solve the problem of using a global synchronous federated computing strategy in the prior art, which requires more manpower and material resources to unify edge devices Management, inefficiency, differences in image angles, etc. lead to inaccurate training

Images