Wyniki wyszukiwania - Biblioteka Nauki

1

Audio-Visual Speech Processing System for Polish Applicable to Human-Computer Interaction

100%

Jadczyk T.

|

2018

|

tom Vol. 19 (1)

41--63

EN

This paper describes audio-visual speech recognition system for Polish language and a set of performance tests under various acoustic conditions. We first present the overall structure of AVASR systems with three main areas: audio features extraction, visual features extraction and subsequently, audiovisual speech integration. We present MFCC features for audio stream with standard HMM modeling technique, then we describe appearance and shape based visual features. Subsequently we present two feature integration techniques, feature concatenation and model fusion. We also discuss the results of a set of experiments conducted to select best system setup for Polish, under noisy audio conditions. Experiments are simulating human-computer interaction in computer control case with voice commands in difficult audio environments. With Active Appearance Model (AAM) and multistream Hidden Markov Model (HMM) we can improve system accuracy by reducing Word Error Rate for more than 30%, comparing to audio-only speech recognition, when Signal-to-Noise Ratio goes down to 0dB.

2

Baza danych słownika języka polskiego ze statystykami słów dla systemu automatycznego rozpoznawania mowy

51%

Mąsior M. , Ziółko B. , Skurzok D. , Jadczyk T.

|

tom Vol. 32, nr 2B

349-357

PL

Artykuł opisuje słownik języka polskiego zaimplementowany w postaci bazy danych na potrzeby systemu rozpoznawania mowy. Przedstawiono zastosowania słownika do poprawienia jakości rozpoznania przez modelowanie języka z wykorzystaniem danych przechowywanych w bazie. Zawarto także informacje na temat danych znajdujących się w bazie na koniec stycznia 2011 roku.

EN

A dictionary of Polish implemented as a data base for automatic speech recognition is presented. The dictionary allows improvement of recognition by language modelling using statistics stored in the data base. The data currently kept in the database are presented as well.

3

Measurement of hydrophobicity distribution in proteins – non-redundant Protein Data Bank

51%

Sałapa K. , Kalinowska B. , Jadczyk T. , Roterman I.

|

tom Vol. 8, no. 3

327--337

EN

The cluster analysis is applied to the analysis of the data describing the status of protein structure in respect to hydrophobic core characteristics. The analysis revealed presence of two clusters distinguishing the proteins accordant with the “fuzzy oil drop” model and those which appear as discordant in respect to this model. The analysis was performed separately for chains treated as structural unit and for units defined according to IV-order (taking the functional protein complex). The characteristics of these two classification system appeared to differ in respect to number of proteins belonging to each of two clusters as well as relation between them.

4

Measurement of hydrophobicity distribution in proteins – complete redundant Protein Data Bank

51%

Sałapa K. , Kalinowska B. , Jadczyk T. , Roterman I.

|

2012

|

tom Vol. 8, no. 2

195--206

EN

The divergence entropy: O/T and O/R measuring the distance between observed/theoretical and observed/random distributions was applied to identify the category of protein structures in respect to the hydrophobic core in protein molecules. The naive interpretation was applied treating the proteins of O/T < O/R as the molecules of hydrophobic core accordant with the theoretically assumed. The proteins of O/T > O/R are treated as representing the hydrophobic core not accordant with the assumed one. The large scale computing was performed (PDB data set) to reveal whether other than simple inequality relation should be used for this identification. The cluster analysis was applied to identify the relation O/T versus O/R as the discrimination factor to classify the category of proteins in respect to their structural form of hydrophobic core.

5

Audiowizualna baza nagrań mowy polskiej

51%

Igras M. , Ziółko B. , Jadczyk T.

Studia Informatica

|

2012

|

tom Vol. 33, nr 2B

163-172

PL

Autorzy prezentują największą, audiowizualną bazę danych mowy polskiej i zarazem jedyną zrealizowaną w jakości HD. Artykuł przedstawia krótki opis podobnych baz dla innych języków oraz opis techniczny wykonanej bazy. Omówiono także napotkane wyzwania w trakcie realizacji bazy danych i jej planowane zastosowania.

EN

The biggest audiovisual database of Polish speech (and the only one made in HD quality) is presented. The paper shortly introduces description of similar databases for other languages and the technical specification of the AGH database. The challenges met during the process of building the database are discussed along with the planned applications.

6

Comparative study of SQLite and Berkeley DB implementations of n-gram model of polish language

45%

Skurzok D. , Ziółko B. , Pohl A. , Jadczyk T. , Mąsior M.

Studia Informatica

|

2012

|

tom Vol. 33, nr 2B

153-162

EN

Aspects of applying databases in computational linguistics are presented. An example of a dictionary and an n-gram model of the AGH automatic speech recognition system is depicted as well. An advantage of Berkeley DB, comparing to SQLite in time efficiency aspect is shown on this case.

PL

Przedstawiono zagadnienia dotyczące stosowania baz danych w lingwistyce komputerowej. Omówiono także przykład słownika i modelu n-gramowego systemu rozpoznawania mowy AGH. Pokazano na tym przykładzie znaczącą przewagę implementacji wykonanej w Berkeley DB nad implementacją SQLite w sensie wydajności czasowej.

7

Computations of protein hydrophobicity profile as virtual experiment in gridspace virtual laboratory

32%

Ciepiela E. , Jadczyk T. , Harężlak D. , Kasztelnik M. , Nowakowski P. , Dyk G. , Malawski M. , Bubak M. , Roterman I.

|

2012

|

tom Vol. 8, no. 4

361--372

EN

The terms like e-science, e-poster or e-health are nowadays commonly used. Special disciplines allowing fast development in these fields of science are commonly available. This paper presents e-paper [1] powered by the Collage Authoring Environment [2] e-publication system which is backed by the GridSpace2 [3] distributed computing platform. This e-publication in a form of WWW page, apart from the traditional textual and graphical content, embeds an on-line software tool for the analysis of the 3-D structure of protein based on the hydrophobicity distribution in protein body. The tool uses GridSpace2 platform in order to carry out computations on the PL-Grid [4] high-performance computing infrastructure. This work shows how this specific e-publication was accomplished utilizing above mentioned already existing information technologies and e-infrastructure. The tool employs the model called “fuzzy oil drop” that assumes the hydrophobicity distribution in proteins being in form of 3-D Gauss function. The protein of the hydrophobicity core structure accordant with the model with all hydrophobic residues buried in the central part of the protein body and hydrophilic residues exposed toward the water environment could be the protein very well soluble although representing no any form of activity. This is why the observed discrepancies between idealized and observed hydrophobicity distribution is presented in form of profile revealing the localization of residues representing local hydrophobicity excess as well as local hydrophobicity deficiency. The distribution of these discrepancies appeared to be specific and function related. The e-publication makes available the tool to calculate the profile of any protein under consideration. The interpretation of the final results is specific for particular protein.