Intelligent Sensory Information Systems
The search engine is fed a huge quantity of image fragments that can be linked to a specific search query. The search engine then assesses each image in colour, texture and shape. Based on this analysis, the search engine will determine a characteristic correlation between the specific distinguishing features and the search query entered by the user. This is in contrast with, for example, Google that relies on users to say what’s in a video or image and is still work in progress. It’s fairly easy to identify a boat but it’s a lot more difficult to identify terms such as ‘love’ or ‘romance’. Or consider filtering out offside in a soccer game?
Arnold Smeulders’ group considers visual information in its many aspects. Research ranges from the perceptual and cognitive processes involved in understanding visual information, methods for learning and indexing the semantics of visual information automatically, methods for the presentation of and interaction with large visual collections and computer vision methods for human computer interaction.