Title: Multilanguage Querying for Image Retrieval

Year of Publication: Nov - 2014
Page Numbers: 21-27
Authors: Arinori Takahashi , Reiko Hishiyama
Conference Name: The International Conference on Computer Graphics, Multimedia and Image Processing (CGMIP2014)
- Malaysia

Abstract:


Images have become important forms of multimedia content on the Internet due to the increased development of multimedia technology. However, image contents cannot be described adequately using tags alone. In addition, the content descriptions are provided in multiple languages and this may make it difficult to acquire appropriate content using a single language. Users may also find it problematic representing their query as text. These issues make it difficult for users to create queries during image retrieval. In this study, we propose a model to facilitate the acquisition of more appropriate images by employing a multilingual query. This model increases the number of initial image candidates by translating the user’s query into the local languages of the retrieval targets and adding the translated query to the original query, thereby increasing the tags searched. Our experimental evaluations showed that users could obtain more appropriate images with this method.