Posted: 03/04/17
A new technology developed by ADAPT Centre researchers based at the School of Computer Science and Statistics at Trinity College Dublin has the ability to provide 3D rendering of people’s sentiments and visual interests. The research, which is funded as part of the FP7 European project GRAISearch, investigates 3D scene rendering with game engine technologies, using information extracted from geo-tagged and up-to-date multimedia data available online.
Speaking about the project Professor Rozenn Dahyot, Project Coordinator, said: “The end goal of this project is a live geo-located social media search engine. One application for this technology is to assist tourists to identify landmarks on 2D or 3D maps when visiting an area. The analysis of social media data relevant to a particular place and time can highlight events and exhibitions, and inform people of other visitors sentiment and experience. The technology integrates graphics with Artificial Intelligence machine learning algorithms to provide capabilities for visualising and navigating in a 3D city generated automatically from open sources such as OpenStreetMap. People’s semntiments and visual interests can then be rendered in 3D using geo-tagged text messages and images posted on social media.”
An example of how the technology works shows a 3D rendering of Trinity College Dublin’s campus in the game engine using spot lights to highlight buildings that appear frequently in images posted on social media. The popular landmarks such as the Campanile light up highlighting the popularity of posting of that image in social media by multiple users.
The innovative work has recently been published in two important journals in the field, Computers & Graphics (DOI:10.1016/j.cag.2017.01.005) and Computer Animation & Virtual World (DOI:10.1002/cav.1742). The work was carried out by Dr Abdullah Bulbul, Marie Curie Fellow on the GRAISearch project, and Project Coordinator, Professor Rozenn Dahyot (Contact: [email protected]). Project GRAISearch (2014-18) is an Industry Academia Partnerships and Pathways European project in collaboration with the Dublin based start-up Tapastreet.
Share this article:
