Cross-modal hash retrieval method based on mapping dictionary learning

Abstract

The invention discloses a mapping dictionary learning based cross-modality hashing search method. The mapping dictionary learning based cross-modality hashing search method includes: collecting image samples and text samples through a network, establishing a cross-media search data set, and dividing the data set into a training set and a test set; extracting features of the images and the texts through a BOW algorithm; learning a sharing sub space for an image mode and a text mode through mapping dictionary learning, and learning hashing functions of each mode at the same time; minimizing quantization errors by learning an orthogonal rotation matrix; calculating hashing codes of test samples through the hashing functions of the image mode and the text mode and the orthogonal rotation matrix; and using test samples of one mode as query samples, using the training set of the other mode as a searched data set, calculating a Hamming distance between the query samples and the searched samples and performing sequencing, and returning samples on top. The search method is high in search accurate rate, can be easily applied to large-scale data sets, is simple to implement, and has wide application prospect and huge market value.

Description

Technical field: [0001] The invention relates to a cross-modal hash retrieval method, in particular to a cross-modal hash retrieval method based on mapping dictionary learning. Background technique: [0002] With the rapid development of computer networks and information technology, the amount of media data on the network has increased dramatically, and the representation of media has also shown multi-modality (image, text, sound, video, etc.). For example: when uploading a photo on Weibo, a paragraph of text is usually uploaded to describe the content of the photo or the content of the image is marked with some tags; When describing product information, both pictures and words are usually used. Although these multimodal data have different representation forms, there are semantic associations between them. The purpose of cross-media retrieval is to mine the semantic relationship between different media, sort according to the semantic relationship, and return data of diffe...

AI Technical Summary

Problems solved by technology

However, such algorithms usually have the following problems, which limit the application of the algorithm

1) In the dictionary learning algorithm, due to the existence of sparse constraint items, the complexity of the training and testing process algorithm is high

2) These hash algorithms do not learn hash functions for each modality

3) The sample representation is sparse, resulting in uneven distribution of -1 and 1 of the hash code

However, these algorithms do not consider the impact of quantization loss on algorithm performance

Images