Gender-dependent (male/female) acoustic models are more acoustically homogeneous and therefore give better recognition performance than single gender-independent model. This paper deals with a problem how to use these gender-based acoustic models in a real-time LVCSR (Large Vocabulary Continuous Speec...

It is well known that gender-dependent (male/female) acoustic models are more acoustically homogeneous and therefore give better recognition performance than single gender-independent model in the case where the gender is successfully detected or a priory known. Speakers do not need to ...

This paper deals with several cohort methods for score normalization in speaker verification systems. At first, the reasons for score normalization are provided. Next, the principle of score normalization techniques based on Bayesian theorem are explained. A new normalization technique, unconstra...

The article introduces an expert system for the speaker veri fcation task. Our main purpose was to design a tool for the combination of various speaker verifcation systems proposed for various operating conditions. First of all, the essential ideas are explained that made us d...

This paper deals with a combination of basic adaptation techniques of Hidden Markov Model used in the speech recognition. The adaptation methods approach the data only through their statistics, which have to be accumulated before the adaptation process. When performing two adaptations s...

This paper deals with robust estimations of data statistics used for the adaptation. The statistics are accumulated before the adaptation process from available adaptation data. In general, only small amount of adaptation data is assumed. These data are often corrupted by noise, channel...

One of the most utilized adaptation techniques is the feature Maximum Likelihood Linear Regression (fMLLR). In comparison with other adaptation methods the number of free parameters to be estimated significantly decreases. Thus, the method is well suited for situations with small amount ...

The aim of this work is to propose a refinement of the shift-MLLR (shift Maximum Likelihood Linear Regression) adaptation of an acoustics model in the case of limited amount of adaptation data, which can lead to ill-conditioned transformations matrices. We try to suppress the ...

In this work we deal with the problem of small amount of data when estimating a feature transformation for the speaker adaptation of an acoustic model. Our goal is to compensate for the lack of adaptation data by a proper initialization of transformation matrices. Methods ...

In the present paper, several experiments on text-to-speech system personification are described. The personification enables TTS system to produce new voices by employing voice conversion methods. The baseline speech synthetizer is a concatenative corpus-based TTS system which utilizes t...

The paper deals with the problem of speaker-based segmentation. The goal of this task is to extract homogeneous segments containing the longest possible utterances produced by a single speaker. In the method presented here, no assumption is made about prior knowledge of the speaker...

The paper deals with the problem of automatic speaker change detection. A metric-based algorithm, called MDISTBIC, which means Modified DISTBIC, is proposed in this paper. The algorithm originates from the DISTBIC algorithm and modifies it in order to reach a higher efficiency. Both...

The paper deals with the problem of automatic speaker change detection. A new metric-based algorithm, called AlgBICMap algorithm, is proposed in this paper. The AlgBICMap algorithm allows to create a map of BIC (Bayessian Information Criterion), which enables us to detect e...

This paper deals with speaker adaptation techniques well suited for the task of online subtitling. Two methods are briefly discussed, namely MAP adaptation and fMLLR. The main emphasis is laid on the description of improvements involved in the process of adaptation subject to the&#...