Method for semantically annotating images on basis of hybrid generative and discriminative learning models

Abstract

The invention discloses a method for semantically annotating images on the basis of hybrid generative and discriminative learning models. The method includes generatively building models of the images by means of continuous PLSA (probabilistic latent semantic analysis) at generative learning stages, acquiring corresponding model parameters and subject distribution of each image, and utilizing the corresponding subject distribution as an intermediate representation vector of each image; constructing cluster classifier chains to discriminatively learn from the intermediate representation vectors of the images at discriminative learning stages, creating the classifier chains and integrating contextual information among annotation keywords; automatically extracting visual features of each given unknown image at annotation stages, acquiring representation of subject vectors of the given unknown images by the aid of a continuous PLSA parameter estimation algorithm, classifying the subject vectors by the aid of trained cluster classifier chains and semantically annotating the images by a plurality of semantic keywords with the highest confidence. The method has the advantage that the annotation and retrieval performance of the method are superior to the annotation and retrieval performance of most current typical methods for automatically annotating images.

Description

technical field [0001] The invention relates to the field of image retrieval, in particular to an image semantic labeling method of a hybrid generative and discriminative learning model. Background technique [0002] According to the characteristics of the machine learning methods used, the existing automatic image labeling methods can be roughly divided into labeling methods based on generative models and labeling methods based on discriminative models. [0003] The characteristics of the annotation method based on the generative model are: first learn the joint probability of image features and keywords, and then calculate the posterior probability of each keyword when the image characteristics are given by Bayesian formula, and perform image annotation according to the posterior probability . Such methods have a scalable training process and require less quality of manual annotation of the training image set. [0004] The characteristics of the labeling method based on ...

AI Technical Summary

Problems solved by technology

This type of method treats each semantic concept as an independent category, and generally can achieve high labeling accuracy, but it is not easy to use domain-related prior knowledge

Images