Structure sparsification maintenance based semi-supervised dictionary learning method

Abstract

The invention discloses a structure sparsification maintenance based semi-supervised dictionary learning method. The method mainly comprises the following steps of firstly establishing a new semi-supervised dictionary learning model through a self-representation relationship between structure sparsification maintenance codes; secondly performing iterative optimization on various variables in the proposed semi-supervised dictionary learning model by adopting a block-coordinate descent method and proving convergence of an algorithm theoretically; and finally proposing a method for constructing class-related sub-dictionaries, and classifying samples through reconstruction errors of the samples in the various sub-dictionaries. According to the method, structure sparsification constraints are introduced, so that a large amount of unlabelled samples can be automatically added into a class in which the unlabelled samples are arranged; and the unlabelled samples and labeled samples in the same class together participate in dictionary learning, so that the sparse representation ability and judgment ability of a dictionary are improved. An experimental result shows that compared with other classic dictionary learning methods, the semi-supervised dictionary learning method has higher classification accuracy, thereby having a very good application prospect.

Description

technical field [0001] The invention relates to the field of computer pattern recognition, in particular to a semi-supervised dictionary learning method based on structural sparsity. Background technique [0002] Dictionary learning is to use training samples to learn an over-complete basis vector set (called dictionary), so as to obtain the sparse representation of input samples under this dictionary. It is one of the research hotspots in the field of computer pattern recognition, and it is widely used in the fields of image denoising, image restoration, image classification and compressed imaging. In general, the current dictionary learning algorithm mainly solves two key problems: (1) the sparse expression ability of the dictionary; (2) the discriminative ability of the dictionary. [0003] Generally speaking, the sparse expressiveness of a dictionary is the ability to accurately reconstruct an input sample with as few atoms as possible in the dictionary. Wright et al. ...

AI Technical Summary

Problems solved by technology

However, the calibration of sample categories is very energy-intensive, and it is very difficult to obtain a large number of labeled samples

Images