We develop an image superresolution method that can the pixel i if zi =-1 due to missing observations, whereas deal with spatially structured noise added to an original image. ordinary small observation noise due to the physical imaging Such a structured noise process can be understood as a model for process is imposed if zi = +1. Thus the hidden variables z possible occlusions such as clouds in the sky or stains on the lens and is modeled as spin glasses. The original high-resolution image rsentm askingomssing-value patter. Wegwilltuse th underlying multiple low-resolution observed images and the hidden Ising model of spin glasses [3] for the spatal organizaton of noise structure are estimated via a variational learning algorithm. noise patterns. Experiments show that our superresolution method can outperform Traditionally, spin glasses have been used for modeling imother methods that do not assume structured noise. other_methods_that do not assume structured noise.ages themselves [4]-[6]. Early studies with a spin-glass image Keywords Image superresolution, structured noise processes, model dealt with artificial images, which typically consist of a spin glasses, the Ising model, variational learning, small number of pixel values, and then, monochrome objects

No abstract

No abstract

SUMMARYIn an image retrieval system, accurately extracting the content of the image which users want to search for based on an image in their heads is vital. As a result, many attempts have been made to use an image similar to the sought-for image as a retrieval key and thereby to increase the freedom of expression for users images. Because of the high variability of image characteristics, however, the vagueness of the content which indicates the retrieval key itself rises compared to conventional methods that use keywords or other linguistic resources for the retrieval key. Therefore, in this research the authors consider these problems by reviewing image retrieval from the perspective of image communications between users and the system, and then propose an interactive mechanism which limits the increase in vagueness while using an image as the retrieval key. The authors then perform retrievals by having users revise the retrieval key based on this mechanism. The authors create their interactive mechanism in a real image retrieval system and confirm its validity through experiments. © 2001 Scripta Technica, Syst Comp Jpn, 32(13): 4151, 2001

We propose a visual query method for image retrieval, in which the user expresses the composition ofthe target image by selecting one of the composition types presented bjl the system, arid also propose an image classijication method to derive composition types from an image database. From the viewpoint of communication between the system and its user, we point out the problem of visual query methods using concrete images as queries. Though suited to convey visual image properties, they grow the ambiguity in the system S interpretation of queries. To manage this trade08 we implemented our methods in a prototype system and derived composition types from an image database to show how our query method works in image retrieval.