Representing data as points in a high-dimensional space, so as to use geometric methods for indexing, is an algorithmic technique with a wide array of uses. It is central to a number of areas such as information retrieval, pattern recognition, and statistical data analysis; many of the problems arising in these applications can involve several hundred or several thousand dimensions. We consider the nearest-neighbor problem for d-dimensional Euclidean space: we wish to pre-process a database of n points so that given a query point, one can efficiently determine its nearest neighbors in the database. There is a large literature on algorithms for this problem, in both the exact and approximate cases. The more sophisticated algorithms typically achieve a query time that is logarithmic in n at the expense of an exponential dependence on the dimension d; indeed, even the averagecase analysis of heuristics such as k-d trees reveals an exponential dependence on d in the query time. In this wor...

The proliferation of multimedia systems is leading to the creation of large on-line collections of visual and multimedia information. A fundamental task consists of locating specific video clips by their content; to obtain this goal a prior video features extraction is needed. The first step involved in the video features extraction task consists of segmenting videos into short sequences called shots. A method able to detect if a scene cut occurs is needed. In this paper a neural architecture for shot detection is introduced. The architecture implements a three-layer Multilayer Perceptron: the learning algorithm is the backpropagation one, and Powells conjugate gradient descent minimisation algorithm is used. Our method is based on the gray-level map of images only, no color information is required. The described technique has been successfully applied to a variety of image sequences with excellent results even in presence of very noisy images. Results on a few videos containing some t...

The focus of our work is the development of a general, scalable architecture to support fast querying of very large image databases with user-specified distance measures. We have developed algorithms and data structures for efficient image retrieval from large databases with multiple distance measures. We are investigating methods for merging our general, distance-measure-independent method with other useful techniques that may be distance measure specific, such as keyword retrieval and relational indexing. We are developing both new methods for combining distance measures and a framework in which users can specify their queries without detailed knowledge of the underlying metrics. We have built a prototype system