Semi-supervision-based central product quantitative retrieval method

Abstract

The invention provides a semi-supervision-based center product quantization image retrieval method, which comprises the following steps of: segmenting a space into a plurality of subspaces after feature extraction, then carrying out normalization processing on a feature vector, and calculating a cosine distance to enable the subvector to find the closest code word in the corresponding subspace. In the calculation process, a semi-supervised loss module is used for reducing quantization errors, minimizing empirical errors of marked data and embedding errors of unmarked data, finally, code words are used for replacing sub-vectorization to form binary codes, the binary codes are stored in a product quantization lookup table, and image retrieval is carried out through asymmetric distance calculation; the method provided by the invention has higher robustness, and the image retrieval precision is improved.

Description

technical field [0001] The invention relates to the technical field of image retrieval, in particular to a semi-supervised central product quantitative retrieval method. Background technique [0002] With the rapid development of the Internet and the popularization of mobile devices, the data accumulated in all walks of life has shown an explosive growth trend. In this era, how users can efficiently and accurately retrieve the required image data information from large-scale image data has become a research hotspot. At present, many scholars have developed fast and accurate retrieval algorithms. Approximate Nearest Neighbor (ANN) has become the most widely used technology in data retrieval due to its high computational speed and retrieval accuracy. [0003] The hash method has obvious advantages in terms of memory consumption and retrieval speed. Specifically, this method maps high-dimensional images into fixed-length hash values ​​and maintains the similarity in the origin...

AI Technical Summary

Problems solved by technology

Most of the existing unsupervised hashing methods adopt a graph-based paradigm, which usually has a "static graph" problem, and most of them only design loss functions to maintain semantic information, and do not take into account the underlying data structure of unlabeled data

Images