SEMI-SUPERVISED IMAGE RECOGNITION AND RETRIEVAL DEVICE, SEMI-SUPERVISED IMAGE RECOGNITION AND RETRIEVAL METHOD, AND PROGRAM

摘要

PROBLEM TO BE SOLVED: To more efficiently learn relationship between an image and language information even when only a few images associated with the language information can be used.;SOLUTION: An accumulation image feature extraction part 1 extracts complete accumulation image features and incomplete accumulation image features from a complete/incomplete accumulation image set. An accumulation additional information feature extraction part 2 extracts accumulation additional information features from the accumulation additional information set. An accumulation latent variable extraction part 3 extracts an accumulation latent variable set as a set of variables for describing relationship between the image and the additional information from the complete/incomplete accumulation image feature set and the accumulation additional information feature set. An inter-image/additional information relationship learning part 4 learns an image/additional information relationship model as a model for describing the relationship between the image and the additional information from the complete/incomplete accumulation image feature set and the accumulation additional information feature set and the accumulation latent variable set.;COPYRIGHT: (C)2011,JPO&INPIT