We address the problem of identifying specific instances of a class (cars) from a set of images all belonging to that class. Although we cannot build a model for any particular instance (as we may be provided with only one "training" example of it), we can use information extracted from observ- ing other members of the class. We pose this task as a learning problem, in which the learner is given image pairs, labeled as matching or not, and must discover which image features are most consistent for matching in- stances and discriminative for mismatches. We explore a patch based representation, where we model the distributions of similarity measure- ments defined on the patches. Finally, we describe an algorithm that selects the most salient patches based on a mutual information criterion.

The Japanese fixed-property tax is imposed by municipalities on the owners of land, buildings, and depreciation assets (all hereinafter referred to as "fixed assets") on January 1 of every year by calculating the tax sum according to current asset values. This identification work is contracted out to survey companies. The identification of such en over a scale that can cover an actual area of 800 changes is entrusted to survey companies who hire by 600 meters or 500 by 600 meters (variable a large number of workers (figure 1, left). However, depending on the municipality), and every municipality reliance on human labor has led to problems has several hundred photographs that must detailed in the following paragraphs. Under these circumstances, the incentives for It takes about 10 hours to read and interpret a single the municipalities to overcome such challenges by photograph, and the average municipality automating or systematizing the photograph-reading must perform this work for several hundred photographs.

The Tokyo Metropolitan Government is the largest municipality in Japan and conducts building change identification work. Recently, Tokyo terminated its traditional visual identification work that has been in use for 20 years and shifted to a new automated system. This paper is intended to introduce the Fixed Assets Change Judgment (FACJ) system and its core tool, RealScape. RealScape automatically detects the changes in the height and color of buildings based on three-dimensional (3D) analysis of aerial photographs. It employs a unique pixel-by-pixel stereo processing method and enables the foot-level precision for each building. RealScape detects building changes more accurately than visual judgment operations by humans and reduces the labor costs to one third of the traditional approach and the required judgment duration to about two weeks per 100km2.

We address the problem of identifying specific instances of a class (cars) from a set of images all belonging to that class. Although we cannot build a model for any particular instance (as we may be provided with only one "training" example of it), we can use information extracted from observing other members of the class. We pose this task as a learning problem, in which the learner is given image pairs, labeled as matching or not, and must discover which image features are most consistent for matching instances and discriminative for mismatches. We explore a patch based representation, where we model the distributions of similarity measurements defined on the patches. Finally, we describe an algorithm that selects the most salient patches based on a mutual information criterion. This algorithm performs identification well for our challenging dataset of car images, after matching only a few, well chosen patches.

We address the problem of identifying specific instances of a class (cars) from a set of images all belonging to that class. Although we cannot build a model for any particular instance (as we may be provided with only one "training" example of it), we can use information extracted from observing other members of the class. We pose this task as a learning problem, in which the learner is given image pairs, labeled as matching or not, and must discover which image features are most consistent for matching instances and discriminative for mismatches. We explore a patch based representation, where we model the distributions of similarity measurements defined on the patches. Finally, we describe an algorithm that selects the most salient patches based on a mutual information criterion. This algorithm performs identification well for our challenging dataset of car images, after matching only a few, well chosen patches.

We address the problem of identifying specific instances of a class (cars) from a set of images all belonging to that class. Although we cannot build a model for any particular instance (as we may be provided with only one "training" example of it), we can use information extracted from observing othermembers of the class. We pose this task as a learning problem, in which the learner is given image pairs, labeled as matching or not, and must discover which image features are most consistent for matching instances anddiscriminative for mismatches. We explore a patch based representation, where we model the distributions of similarity measurements definedon the patches. Finally, we describe an algorithm that selects the most salient patches based on a mutual information criterion. This algorithm performs identification well for our challenging dataset of car images, after matching only a few, well chosen patches.