Speech compression, enhancement and recognition in noisy, reverberant conditions is a challenging task. In this paper a new approach to this problem, which is developed in the framework of probabilistic random modeling. speech coding techniques are commonly used in low bit rate analysis and synthesis . Coding algorithms seek to minimize the bit rate in the digital representation of a signal wit...

As a major Chinese dialect, Cantonese is well known for its complex tone system. This paper presents a preliminary analysis of the F0 contours of Cantonese using a command-response model. Experiments are conducted on a set of designed sentences, from which a set of appropriate tone command patterns for each tone is derived by Analysis-by-Synthesis. The model shows a high accuracy of approximati...

This paper describes a robust speech documents retrieval system that uses voice input keywords. To solve the inevitable problems which arise when the input to the system is speech, i.e. misrecognition, a novel method was developed, where, before the retrieval processing, unproductive keyword candidates are discarded by a grouping processing using the similarity between words and the recognition...

Speech coding is a field where compression paradigms have not changed in the last 30 years. The speech signals are most commonly encoded with compression methods that have roots in Linear Predictive theory dating back to the early 1940s. This paper tries to bridge this influential theory with recent cognitive studies applicable in speech communication engineering. This tutorial article reviews ...

This paper proposes a speech codec based on the MultiPulse based CELP (MP-CELP) coding and a convolutional coding algorithms for the ETSI Adaptive Multi-Rate (AMR) standard. The codec operates at several speech coding rates, maintaining a fixed gross rate including speech and channel coding for the Full-Rate (FR) and Half-Rate (HR) channel modes. MP-CELP has great features of easily changing th...

This paper describes an algorithm for finding all and only the structural descrip­ tions of any sentence with respect to a given grammar constructed in accordance with the theory of grammar which has been developed by linguists at the Massachusetts Institute of Technology — for the most part by N. Chomsky [1] — and which is discussed by E. S. Klima at this colloquium [2]. In an earlier paper [3...

Decomposition of speech signals into periodic and noise components is widely used in speech coding to facilitate efficient compression. Existing decomposition schemes are either too inflexible to model transient changes in the speech signal and require large delay or alternatively produce a large set of parameters that is not scalable to low rate applications. This paper proposes a scheme that ...

This study offers a phonetic description and phonological analysis of a variety of melodic contours in Brazilian Portuguese. These are commonly referred to as conveying the modal, or illocutionary, functions of intonation, as well as some attitudinal uses. The phonological analysis follows the AM framework and aims at distinguishing what is contrastive and what are merely allophonic variations ...

The number of units in an utterance determines how much time speakers require to physically plan and begin their production [1]-[2]. Previous research proposed that the crucial units are prosodic i.e., Phonological Words (PWs), not syntactic or morphological [3]. Experiments on Dutch using a prepared speech paradigm claimed to support this view [4]-[5]; however, compounds did not conform to pre...

F a c u l t y M e n t o r Cristina Videira Lopes School of Information & Computer Science