TY - CONF
AU - Römke, Thomas
AU - Röttger, Markus
AU - Schroeder, Ulf-Peter
AU - Simon, Jens
ID - 2448
T2 - Proc. ZEUS Workshop on Par. Programming and Computation
TI - An Efficient Mapping Library for Parix
ER -
TY - GEN
AU - Blömer, Johannes
AU - Kalfane, Malik
AU - Karp, Richard
AU - Karpinski, Marek
AU - Luby, Michael
AU - Zuckerman, David
ID - 3038
TI - An XOR-based erasure-resilient coding scheme
ER -
TY - GEN
AU - Peckhaus, Volker
ID - 17776
T2 - Berichte zur Wissenschaftsgeschichte
TI - André Weil, Lehr- und Wanderjahre eines Mathematikers, aus dem Französischen von Theresia Übelhör, Birkhäuser Verlag: Basel/Boston/Berlin 1993
VL - 18
ER -
TY - CONF
AB - Clustering techniques have been integrated at different levels into the training procedure of a continuous-density hidden Markov model (HMM) speech recognizer. These clustering techniques can be used in two ways. First acoustically similar states are tied together. It will help to reduce the number of parameters but also allow to train otherwise rarely seen states together with more robust ones (state-tying). Secondly densities are clustered across states, this reduces the number of densities while at the same time keeping the best performances of our recognizer (density-clustering). We have applied these techniques both to word-based small-vocabulary and phoneme-based large-vocabulary recognition tasks. On the WSJ task, we could achieve a reduction of the word error rate by 7%. On the TI/NIST-connected digit task, the number of parameters was reduced by a factor 2-3 while keeping the same string error rate.
AU - Dugast, Christian
AU - Beyerlein, Peter
AU - Haeb-Umbach, Reinhold
ID - 11757
T2 - ICASSP, Detroit
TI - Application of Clustering Techniques to Mixture Density Modelling for Continuous-Speech Recognition
ER -
TY - JOUR
AB - For two given simple polygonsP, Q, the problem is to determine a rigid motionI ofQ giving the best possible match betweenP andQ, i.e. minimizing the Hausdorff distance betweenP andI(Q). Faster algorithms as the one for the general problem are obtained for special cases, namely thatI is restricted to translations or even to translations only in one specified direction. It turns out that determining pseudo-optimal solutions, i.e. ones that differ from the optimum by just a constant factor, can be done much more efficiently than determining optimal solutions. In the most general case, the algorithm for the pseudo-optimal solution is based on the surprising fact that for the optimal possible match betweenP and an imageI(Q) ofQ, the distance between the centroids of the edges of the convex hulls ofP andI(Q) is a constant multiple of the Hausdorff distance betweenP andI(Q). It is also shown that the Hausdorff distance between two polygons can be determined in timeO(n logn), wheren is the total number of vertices.
AU - Alt, Helmut
AU - Behrends, Bernd
AU - Blömer, Johannes
ID - 3037
IS - 3
JF - Annals of Mathematics and Artificial Intelligence
SN - 1573-7470
TI - Approximate matching of polygonal shapes
VL - 13
ER -
TY - CONF
AB - We present a combination of recent extensions to single-pushout graph transformations, as there are attribution, application conditions and amalgamated graph transformations and add a simple transaction concept on top of this formalism. Thereby, we provide the formal basis for several examples, where these concepts are used in combination.
AU - Heckel, Reiko
AU - Müller, Jürgen
AU - Taentzer, Gabriele
AU - Wagner, Annika
ID - 7831
IS - B-19
T2 - Proceedings of the Colloquium on Graph Transformation and its Application in Computer Science
TI - Attributed Graph Transformations with Controlled Application of Rules
ER -
TY - CONF
AB - We address the problem of automatically finding an acoustic representation (i.e. a transcription) of unknown words as a sequence of subword units, given a few sample utterances of the unknown words, and an inventory of speaker-independent subword units. The problem arises if a user wants to add his own vocabulary to a speaker-independent recognition system simply by speaking the words a few times. Two methods are investigated which are both based on a maximum-likelihood formulation of the problem. The experimental results show that both automatic transcription methods provide a good estimate of the acoustic models of unknown words. The recognition error rates obtained with such models in a speaker-independent recognition task are clearly better than those resulting from separate whole-word models. They are comparable with the performance of transcriptions drawn from a dictionary.
AU - Haeb-Umbach, Reinhold
AU - Beyerlein, P.
AU - Thelen, E.
ID - 11787
T2 - ICASSP, Detroit
TI - Automatic Transcription of Unknown Words in a Speech Recognition System
ER -
TY - JOUR
AU - Hellebrand, Sybille
AU - Rajski, Janusz
AU - Tarnick, Steffen
AU - Venkataraman, Srikanth
AU - Courtois, B.
ID - 13011
IS - 2
JF - {IEEE Transactions on Computers}
TI - Built-In Test for Circuits with Scan Based on Reseeding of Multiple-Polynomial Linear Feedback Shift Registers
VL - 44
ER -
TY - GEN
AU - Eke, Norbert Otto
ID - 19556
IS - H. 1
T2 - Zeitschrift für Germanistik
TI - Bürgerlichkeit im Umbruch. Studien zum deutschsprachigen Drama 1750-1800. Mit einer Bibliographie der Dramen der Oettingen-Wallersteinschen Bibliothek zwischen 1750 und 1800. Hg. von Helmut Koopmann. Tübingen 1993
VL - NF V
ER -
TY - CHAP
AU - Oevel, Gudrun
AU - Siek, G.
ID - 17307
T2 - AACE educational Multimedia and Hypermedia Annual
TI - Computer algebra in education
ER -