Distributed image search method based on supervised learning

Abstract

The invention discloses a distributed image search method based on supervised learning. Firstly, images, videos and files are classified and marked in a database of each node. A classification matrix,a coding matrix, a hash code matrix and corresponding Lagrange multipliers are initialized, then minimum classification errors and reconstruction errors are introduced to construct an objective function. The objective function is solved, and a parameter matrix is updated. Each data node communicates with a central node, whether the conversion matrix of each node tend to be consistent or not is judged, a Lagrange multiplier is uptaded, and finally an approximation search process is performed. The problems that the required scale of large-scale data during storage and calculation is too large,and an algorithm model is trained in a centralized mode and is not suitable any more are solved; and each data node and the central node communicate without exchanging original information, so that the problem of overlarge transmission communication can be effectively solved, and meanwhile, data on each nodes is kept independent.

Description

technical field [0001] The invention relates to an image search method, in particular to a distributed image search method, which belongs to the field of machine learning. Background technique [0002] With the continuous development of social networks, e-commerce, mobile Internet, etc., the scale of data storage and processing is getting larger and larger, and stand-alone systems can no longer meet the growing needs. Internet companies such as Google and Alibaba have successfully spawned the two hot fields of cloud computing and big data. Both cloud computing and big data are applications built on distributed storage. The core of cloud storage is the back-end large-scale distributed storage system. Big data not only needs to store massive amounts of data, but also needs to analyze these data through appropriate frameworks and tools to obtain useful parts. If there is no distributed storage, it will Not to mention the analysis of big data. Although the research on distribu...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a distributed image search method based on supervised learning, which is mainly used to solve the problem that the number of samples such as images, videos, and texts is large, and it is impossible to accurately find semantic neighbors. If they are trained together, the amount of transmission and calculation The main purpose of this method is to obtain a globally optimized coding matrix with low computational overhead through distributed training, while protecting the data independence of each node in distributed training, and realizing the neighbor search of query samples

Images