A cross-modal hash learning method based on anchor graph

Abstract

A cross-modal hash learning method based on anchor graph, which is characterized in that the method includes the following steps: (1) Using an objective function designed based on anchor graph technology to obtain the binary hash of an object in image modality and text modality (2) In view of the non-convex nature of the objective function, the unknown variables , , and in the objective function are solved by alternately updating; (3) Based on the solution, we get The projection matrix sum of the image modalities and text modalities, generate binary hash codes for the query samples and samples in the retrieval sample set; (4) Calculate the Hamming of the query samples to each sample in the retrieval sample set based on the generated binary hash codes distance; (5) The retrieval of query samples is done using a cross-modal retriever based on approximate nearest neighbor search. This method can quickly obtain the approximate matrix of the real similarity matrix based on the anchor graph technology.

Description

technical field [0001] The invention relates to a cross-modal hash learning method based on anchor graph. Background technique [0002] With the rapid development of information technology, human society has entered the era of big data, and massive amounts of data from different fields and applications are generated all the time. Faced with the explosive growth of data, how to quickly retrieve the required information from it, so as to ensure the effective use of the data, has become an urgent and very challenging problem to be solved in the era of big data. [0003] Nearest neighbor search, also known as similarity search, plays an important role in many applications such as document retrieval, object recognition, and approximate image detection. Among many methods for approximate nearest neighbor search, hash-based search (retrieval) methods have received more and more attention in recent years. The hash-based search method can map high-dimensional feature data into a co...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to provide a cross-modal hash learning method based on anchor graphs, which solves the problem that existing cross-modal hash learning methods have not completely solved the problem of maintaining features based on graph structures on large-scale data sets. The problem of beneficial information in the data, and the problem of discriminative feature selection involved in mapping raw feature data from high-dimensional feature space to low-dimensional Hamming space, propose a cross-modal hashing learning method based on anchor graph and applied to cross-modal retrieval tasks involving image modality and text modality

Images