Mental image retrieval is an original interactive search paradigm introduced by IMEDIA to solve the page zero problem (the user does’nt have any explicit query to start the search). In this framework, the user is supposed to only have a visual representation of the targeted content in mind. Search is then performed by user/machine interactions on the visual contents, without any prior information on the user’s perception of similarity.
Some publications: image category search from a mental picture, survey on mental image retrieval [MLTL2008], Interactive Image Retrieval by Mental Matching[MIR2006],[MIR2005]
