Text-to-speech (TTS) is currently a mature technology that is used in many applications. Some modules of a TTS depend on the language and, while there are many public resources for English, the resources for some underrepresented languages are still limited. This work describes the development of a complete TTS system for Brazilian Portuguese which expands the already available resources. The s...

It is well known that frame independence assumption is a fundamental limitation of current HMM based speech recognition systems. By treating each speech frame independently, HMMs fail to capture trajectory information in the acoustic signal. This paper introduces Gaussian Transition Models (GTM) to model trajectories implicitly. Comparing to alternative approaches, such as segment modeling and ...

Good HMM-based speech recognition performance requires at most minimal inaccuracies to be introduced by HMM conditional independence assumptions. In this work, HMM conditional independence assumptions are relaxed in a principled way. For each hidden state value, additional dependencies are added between observation elements to increase both accuracy and discriminability. These additional depend...

Speech recognition is usually based on Hidden Markov Models (HMMs), which represent the temporal dynamics of speech very efficiently, and Gaussian mixture models, which do non-optimally the classification (acoustic modeling) of speech into single speech units (phonemes). In this paper we present an overview about the use of Support Vector Machines (SVMs) for the classification task by integrati...

A generalised fuzzy approach to statistical modelling techniques for speech recognition is proposed in this paper. Fuzzy C-means (FCM) and fuzzy entropy (FE) techniques are combined into a generalised fuzzy technique and applied to hidden Markov models (HMMs). A more robust version of the above fuzzy technique based on the noise clustering (NC) method is also proposed. Experimental results were...

Our recent experiments with HMM-based speech synthesis systems have demonstrated that speaker-adaptive HMM-based speech synthesis (which uses an ‘average voice model’ plus model adaptation) is robust to non-ideal speech data that are recorded under various conditions and with varying microphones, that are not perfectly clean, and/or that lack of phonetic balance. This enables us consider buildi...

In spontaneous speech, various speech style and speed changes can be observed, which are known to degrade speech recognition accuracy. In this paper, we describe an optimized multi-duration HMM (OMD). An OMD is a kind of multi-path HMM with at most two parallel paths. Each path is trained using speech samples with short or long phoneme duration. The thresholds to divide samples of phonemes are ...

This contribution is about experiments in audio-visual isolated words recognition. The results of these experiments will be used to improve our voice dialogue system, where visual speech recognition will be added. The voice dialogue systems can be used in train or bus stations (or elsewhere), where noise levels are relatively high, therefore the visual part of speech can improve the recognition...

This paper describes a technique for synthesizing synchronized lip movements from auditory input speech signal. The technique is based on an algorithm for parameter generation from HMM with dynamic features, which has been successfully applied to text-to-speech synthesis. Audio-visual speech unit HMMs, namely, syllable HMMs are trained with parameter vector sequences that represent both auditor...

Overlapping speech is still a major cause of error in many speech processing applications, currently without any satisfactory solution. This paper considers the problem of detecting segments of overlapping speech within meeting recordings. Using an HMM-based framework recordings are segmented into intervals containing non-speech, speech and overlapping speech. New to this contribution is the us...