Abstract:

A transfer learning based text-image feature mapping algorithm under event constraint is proposed. Firstly, the documents of each event are modeled by the latent dirichlet allocation, in which the most discriminating feature is obtained by computing the information gain of each topic. Secondly, the images of the corresponding event are modeled through the bag-of-visual-word model and the nave bayes approach. Finally, the feature distributions of the target images are approximated by utilizing the feature distributions of the text data and the text-image co-occurrence data within the same event. Experiment is conducted on a dataset containing 15 categories of events. The effectiveness of the proposed feature mapping algorithm is shown.

Key words: event constraint, transfer learning, text-image feature mapping, co-occurrence data