Authors : Ghaidaa A. Al-Sultany and Suha Kamal
Abstract: In this study, the issue of text modeling was considered with an annotated textual description for diverse categories of images. A topic modeling with the Latent Dirichlet Allocation technique were proposed on a set of images text descriptions to realize the precise topic for each set of images with respect to their attached descriptions from which classifying new images based on the observed topics would be more facilitated. The research was evaluated on the benchmark CLEF dataset; the results were encouraging with regards to the enhancing image retrieval using topics extraction.
Ghaidaa A. Al-Sultany and Suha Kamal, 2017. Image Categorization using Topic Modeling with the Latent Dirichlet Allocation. Journal of Engineering and Applied Sciences, 12: 4123-4126.