Svetlana Kordumova’s research focuses on teaching machines to recognise images without needing to provide direct manual annotation data. Instead, Kordumova relies on tagged data from social media platforms for the recognition of concepts or on objects’ semantics and layout to recognise scenes.
|Date||2 November 2016|
S. Kordumova: Learning to Search for Images without Annotations.
Prof. A.W.M. Smeulders
Dr C.G.M. Snoek
This event is open to the public.