Multimodal Photo Retrieval through Finding Similar Documents Enhanced with Visual Clues - a Baseline Method

Image retrieval till today is one of the most challenging problems in computer science. Even though there are lots of researches performed around the World, an efficient, user friendly image retrieval system still seems to be an unachievable goal. ImageCLEF Photo Retrieval Track allows to compare various approaches to this challenging problem. In this paper we present a staring point of our research, connected to a joint Polish–Singaporean research project, titled: Framework for Visual Information Retrieval and Building Content-based Visual Search Engines. Various techniques, published in the literature are gathered and orchestrated together. A reference image retrieval system is build, supporting both image queries, text queries and joint textimage queries. In our work we have tried to capture state-of-the-art in text and image retrieval.