We explore the use of prosodic features beyond pauses, including duration, pitch, and energy features, for automatic sentence segmentation of ICSI meeting data. Results show that (1) information from pauses is important, including pause duration both at the boundary and at the previous&...

Dialog act (DA) segmentation in meeting speech is important for meeting understanding. In this paper, we explore speaker adaptation of hidden event language models (LMs) for DA segmentation using the ICSI Meeting Corpus. Speaker adaptation is performed using a linear combination of the&...

Structural metadata extraction (MDE) research aims to develop techniques for automatic conversion of raw speech recognition output to forms that are more useful to humans and to downstream automatic processes. It may be achieved by inserting boundaries of syntactic/ semantic unit...

Co-channel speech occurs in situations when one speaker's speech is corrupted by another speaker's speech. Such situations occur for example in TV discussions, telephone calls, etc. Since the parts of speech signal containing multiple voices can cause problems in automatic speech or&...

Recent research progress in developing of the Czech – Sign Speech synthesizer is presented. The current goal is to improve the system for automatic synthesis to produce accurate synthesis of the Sign Speech. The synthesis system converts written text to an animation of an arti...

In this paper, we present the setup of a Wizard of Oz environment used for collection of data for the implementation of the Czech Senior Companion dialogue system. We also discuss some aspects of using WoZ method for collection of emotional data and summarize some statist...

Our paper deals with the multi-label text classification of the newspaper articles, where the classifier must decide if a document does or does not belong to each topic from the predefined topic set. A generative classifier is used to tackle this task and the problem with...

The report describes two UWB systems submitted to the EVALITA 2009 evaluation campaign. Both systems are based on the UBM-GMM approach. Our main motivation laid in the investigation of complementarity of simple UBM-GMM systems in order to achieve a robust performance in di erent&#x...

This paper deals with adaptation techniques based on maximum likelihood linear transformations, which are well suited for the task of on-line recognition. When transcriptions are available before the system starts running, we are speaking about supervised adaptation. In unsupervised adaptation th...

This paper investigates the combination of discriminative adaptation techniques. The discriminative Maximum A-Posteriori (DMAP) adaptation and discriminative feature Maximum Likelihood Linear Regression (DfMLLR) are examined. Since each of the methods is proposed for distinct amount of adaptation data it&#...

Gaussian Mixture Model (GMM) statistics are required for maximum likelihood training as well as for adaptation techniques. In order to train/adapt a reliable model a lot of data are needed, what makes the estimation process time consuming. The paper presents an efficient implementa...

In the paper recent methods used in the task of speaker recognition are presented. At first, the extraction of so called i-vectors from GMM based supervectors is discussed. These i-vectors are of low dimension and lie in a subspace denoted as Total Variability Space (TVS)....

In the paper Factor Analysis (FA) and Nuisance Attribute Projection (NAP) are reviewed, analyzed and compared. Since nowadays FA become a part of most state-of-the-art recognition systems (used e.g. in the concept of i-vectors or PLDA models) it is of relevance to examine different...

This paper deals with several adaptation techniques, which are of the importance in cases when the identity of a speaker is known and we want to recognize his speech. Each of the methods yields various benefits, therefore we examined their combination. This approach brought fu...

This paper presents ARTIC, a brand-new Czech text-to-speech (TTS) system. ARTIC (ARtificial Talker In Czech) is a concatenation-based system that consists of three main, relatively independent, components: speech segment database, text analyzer and speech synthesizer. A statistical approach to speech&...

This paper deals with the design of a speech corpus for a concatenation-based text-to-speech (TTS) synthesis. Several aspects of the design process are discussed here. We propose a sentence selection algorithm to choose sentences (from a large text corpus) which will be read and&#x...

This paper presents an experimental German speech synthesis system. As in case of a Czech text-to-speech system ARTIC, statistical approach (using hidden Markov models) was employed to build a speech segment database. This approach was confirmed to be language independent and it was...

This paper deals with the problems of automatic segmentation for the purposes of Czech concatenative speech synthesis. Statistical approach to speech segmentation using HMMs is applied in the baseline system. Several improvements of this system are then proposed to get more accurate seg...

This paper presents recent improvements on ARTIC - the modern Czech corpus-based text-to-speech system. As a statistical approach (using hidden Markov models) was applied to create an acoustic unit inventory, several improvements concerning acoustic unit modelling, clustering and segmentation have...

This paper gives a survey of the current state of ARTIC -- the modern Czech concatenative corpus-based text-to-speech system. All stages of the system design are described in the paper, including the acoustic unit inventory building process, text processing and speech production issues....