Visual Query Posing in Multimedia Web Document Retrieval

• Published in: 2021 IEEE 15th International Conference on Semantic Computing (ICSC) pp. 415 - 420
• Main Authors: Rinaldi, Antonio M., Russo, Cristiano, Tommasino, Cristian
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.01.2021
• Subjects: Content based image retrieval; deep features; Deep learning; Image retrieval; Multimedia computing; multimedia document retrieval; Ontologies; Search engines; Semantic information retrieval; Semantics; Visualization
• DOI: 10.1109/ICSC50631.2021.00086

In the last few years, the techniques implemented to satisfy users' information needs are changing due to the heterogeneity of web contents, which are getting more wealthy of multimedia contents. The search engines adapt at information needs of users and provides a multi-modal search interface that allows the users to specify the image and text predominantly. Nevertheless, it is not easy to find systems that integrate visual and textual query with reasonable accuracy. Over the years, many authors proposed several frameworks that develop and improve search engines. Moreover, the use of ontology in semantic research had good results and, in an image retrieval system, the deep learning obtained excellent outcomes. In this paper, we present a novel approach for visual query posing to perform a visual and textual query using only images.