Multi-modal Indexing and Retrieval Using an LSA-Based Kernel

摘要

This article proposes a Latent Semantic Analysis (LSA) based kernel function which effectively integrates low-level visual features with higher semantic ones into a common latent space that correlates multimodal features in the same latent space. The method's potential was evaluated on two early fusion experiments in a realistic scenario of image retrieval as the one provided by the ImageCLEF medical task. The performance of the method depends on the distributions of the multimodal latent features and the experimental results show that it outperforms the latent indexing generated by singular value decomposition.