Fusing semantic aspects for image annotation and retrieval

In this paper, we present an approach based on probabilistic latent semantic analysis (PLSA) to achieve the task of automatic image annotation and retrieval. In order to model training data precisely, each image is represented as a bag of visual words. Then a probabilistic framework is designed to capture semantic aspects from visual and textual modalities, respectively. Furthermore, an adaptive asymmetric learning algorithm is proposed to fuse these aspects. For each image document, the aspect distributions of different modalities are fused by multiplying different weights, which are determined by the visual representations of images. Consequently, the probabilistic framework can predict semantic annotation precisely for unseen images because it associates visual and textual modalities properly. We compare our approach with several state-of-the-art approaches on a standard Corel dataset. The experimental results show that our approach performs more effectively and accurately.