Recently, research in the field of content-based image retrieval has attracted a lot of attention. Nevertheless, most existing methods cannot be easily applied to medical image databases, as global image descriptions based on color, texture, or shape do not supply sufficient semantics for medical applications. The concept for content-based image retrieval in medical applications (IRMA) is therefore based on the separation of the following processing steps: categorization of the entire image; registration with respect to prototypes; extraction and query-dependent selection of local features; hierarchical blob representation including object identification; and finally, image retrieval. Within the first step of processing, images are classified according to image modality, body orientation, anatomic region, and biological system. The statistical classifier for the anatomic region is based on Gaussian kernel densities within a probabilistic framework for multiobject recognition. Special emphasis is placed on invariance, employing a probabilistic model of variability based on tangent distance and an image distortion model. The performance of the classifier is evaluated using a set of 1617 radiographs from daily routine, where the error rate of 8.0% in this six-class problem is an excellent result, taking into account the difficulty of the task. The computed posterior probabilities are furthermore used in the subsequent steps of the retrieval process.

visualization of the problem to model human speech. A large number of example sequences of observation vectors (shown connected as continuous trajectories) depending on a given sequence of class labels, with each class representing for example a phoneme (here the name Keiko with given durations). In this synthetic example, the one-dimensional target data would be represented poorly by a uni-modal Gaussian distribution with a constant variance (which corresponds to using the squared-error objective function), which would average the two separate branches, indicated by the fat lines as the mean and constant variance of the single Gaussian. Compare this figure with Figure 3.10, Figure 3.11 and Figure 3.12 to see a subsequent improvement of the model.

Normally compressed data needs to be decompressed before it is processed, but if the compression has been done in the fight way, it is often possible to search the data without having to decompress it, or at least only partially decompress it. The problem can be divided into lossless and lossy compression methods, and then in each of these cases the pattern matching can be either exact or inexact. Much work has been reported in the literature on techniques for all of these cases, including algorithms that are suitable for pattern matching for various compression methods, and compression methods designed specifically for pattern matching. This work is surveyed in this paper. The paper also exposes the important relationship between pattern matching and compression, and proposes some performance measures for compressed pattern matching algorithms. Ideas and directions for future work are also described.

Abstract not found

Searching multimedia data in particular audiovisual data is still a challenging task to fulfill. The number of digital video recordings has increased dramatically as recording technology has become more affordable and network infrastructure has become easy enough to provide download and streaming solutions. But, the accessibility and traceability of its content for further use is still rather limited. In our paper we are describing and evaluating a new approach to synchronizing auxiliary text-based material as, e. g. presentation slides with lecture video recordings. Our goal is to show that the tentative transliteration is sufficient for synchronization. Different approaches to synchronize textual material with deficient transliterations of lecture recordings are discussed and evaluated in this paper. Our evaluation data-set is based on different languages and various speakers’ recordings.

In this paper we introduce a well-motivated abstract pruning criterion for LVCSR decoders based on the anticipated recombination of HMM state alignment paths. We show that several heuristical pruning methods common in dynamic network decoders are approximations of this pruning criterion. The abstract criterion is too complex to be applied directly in an efficient manner, so we derive approximations which can be applied efficiently. Our new pruning methods allow much more exhaustive pruning of the search space than previous methods. We show that the size of the search space can be reduced by up to 50 % at equal precision over the previous state of the art, and the RTF by 20%. The abstract pruning criterion can be considered a guide to derive effective pruning methods for any kind of time synchronous decoder. Index Terms: speech recognition, search, pruning 1.