A discriminative and sparse topic model for image classification and annotation

作者:

Highlights:

• The label information is enforced in the generation of visual and annotation terms.

• The zero-mean Laplace distribution is added to a topic generative process.

• The sparse image representation is helpful to learn a training model.

• A series of experiments on four data sets demonstrate the performance of DSTM.

摘要

•The label information is enforced in the generation of visual and annotation terms.•The zero-mean Laplace distribution is added to a topic generative process.•The sparse image representation is helpful to learn a training model.•A series of experiments on four data sets demonstrate the performance of DSTM.

论文关键词:Graphical model,Discriminative topic,Sparsity,Image classification,Image annotation

论文评审过程:Received 14 June 2014, Revised 22 August 2015, Accepted 16 March 2016, Available online 5 April 2016, Version of Record 24 April 2016.

论文官网地址:https://doi.org/10.1016/j.imavis.2016.03.005