Search results for "Speech processing"
showing 10 items of 210 documents
On the empirical spectral distribution for certain models related to sample covariance matrices with different correlations
2021
Given [Formula: see text], we study two classes of large random matrices of the form [Formula: see text] where for every [Formula: see text], [Formula: see text] are iid copies of a random variable [Formula: see text], [Formula: see text], [Formula: see text] are two (not necessarily independent) sets of independent random vectors having different covariance matrices and generating well concentrated bilinear forms. We consider two main asymptotic regimes as [Formula: see text]: a standard one, where [Formula: see text], and a slightly modified one, where [Formula: see text] and [Formula: see text] while [Formula: see text] for some [Formula: see text]. Assuming that vectors [Formula: see t…
Probabilities to Accept Languages by Quantum Finite Automata
1999
We construct a hierarchy of regular languages such that the current language in the hierarchy can be accepted by 1-way quantum finite automata with a probability smaller than the corresponding probability for the preceding language in the hierarchy. These probabilities converge to 1/2.
ON-LINE CONSTRUCTION OF A SMALL AUTOMATON FOR A FINITE SET OF WORDS
2012
In this paper we describe a "light" algorithm for the on-line construction of a small automaton recognising a finite set of words. The algorithm runs in linear time. We carried out good experimental results on real dictionaries, on biological sequences and on the sets of suffixes (resp. factors) of a set of words that shows how our automaton is near to the minimal one. For the suffixes of a text, we propose a modified construction that leads to an even smaller automaton. We moreover construct linear algorithms for the insertion and deletion of a word in a finite set, directly from the constructed automaton.
Automata and forbidden words
1998
Abstract Let L ( M ) be the (factorial) language avoiding a given anti-factorial language M . We design an automaton accepting L ( M ) and built from the language M . The construction is effective if M is finite. If M is the set of minimal forbidden words of a single word ν, the automaton turns out to be the factor automaton of ν (the minimal automaton accepting the set of factors of ν). We also give an algorithm that builds the trie of M from the factor automaton of a single word. It yields a nontrivial upper bound on the number of minimal forbidden words of a word.
Tag2Risk: Harnessing social music tags for characterizing depression risk
2020
Musical preferences have been considered a mirror of the self. In this age of Big Data, online music streaming services allow us to capture ecologically valid music listening behavior and provide a rich source of information to identify several user-specific aspects. Studies have shown musical engagement to be an indirect representation of internal states including internalized symptomatology and depression. The current study aims at unearthing patterns and trends in the individuals at risk for depression as it manifests in naturally occurring music listening behavior. Mental well-being scores, musical engagement measures, and listening histories of Last.fm users (N=541) were acquired. Soci…
Low-Power Audio Keyword Spotting using Tsetlin Machines
2021
The emergence of Artificial Intelligence (AI) driven Keyword Spotting (KWS) technologies has revolutionized human to machine interaction. Yet, the challenge of end-to-end energy efficiency, memory footprint and system complexity of current Neural Network (NN) powered AI-KWS pipelines has remained ever present. This paper evaluates KWS utilizing a learning automata powered machine learning algorithm called the Tsetlin Machine (TM). Through significant reduction in parameter requirements and choosing logic over arithmetic based processing, the TM offers new opportunities for low-power KWS while maintaining high learning efficacy. In this paper we explore a TM based keyword spotting (KWS) pipe…
Determination of m¯b/m¯c and m¯b from nf=4 lattice QCD+QED
2021
We extend HPQCD's earlier ${n}_{f}=2+1+1$ lattice-QCD analysis of the ratio of $\overline{\mathrm{MS}}$ masses of the $b$ and $c$ quark to include results from finer lattices (down to 0.03 fm) and a new calculation of QED contributions to the mass ratio. We find that ${\overline{m}}_{b}(\ensuremath{\mu})/{\overline{m}}_{c}(\ensuremath{\mu})=4.586(12)$ at renormalization scale $\ensuremath{\mu}=3\text{ }\text{ }\mathrm{GeV}$. This result is nonperturbative. Combining it with HPQCD's recent lattice $\mathrm{QCD}+\mathrm{QED}$ determination of ${\overline{m}}_{c}(3\text{ }\text{ }\mathrm{GeV})$ gives a new value for the $b$-quark mass: ${\overline{m}}_{b}(3\text{ }\text{ }\mathrm{GeV})=4.513(2…
Languages with mismatches
2007
AbstractIn this paper we study some combinatorial properties of a class of languages that represent sets of words occurring in a text S up to some errors. More precisely, we consider sets of words that occur in a text S with k mismatches in any window of size r. The study of this class of languages mainly focuses both on a parameter, called repetition index, and on the set of the minimal forbidden words of the language of factors of S with errors. The repetition index of a string S is defined as the smallest integer such that all strings of this length occur at most in a unique position of the text S up to errors. We prove that there is a strong relation between the repetition index of S an…
Linguistic interpretation of speech errors
2016
The paper is an attempt to illustrate the linguistic interpretation of speech, known that it remains insufficiently resolved, especially for Romanian. The cause is given by the multitude of criteria that can or should be considered important in speech processing. The aim of this study is to develope a computational tool in order to identify the possible errors related to the morphosintactic structure of speech. Our goal is to assist users who can receive automatically different suggestions that can help them to improve the quality of their text. Thus, we chose an interdisciplinary approach through speech analysis that brings together the key fields of linguistics, computer science and so on…
A practical solution to the problem of automatic word sense induction
2004
Recent studies in word sense induction are based on clustering global co-occurrence vectors, i.e. vectors that reflect the overall behavior of a word in a corpus. If a word is semantically ambiguous, this means that these vectors are mixtures of all its senses. Inducing a word's senses therefore involves the difficult problem of recovering the sense vectors from the mixtures. In this paper we argue that the demixing problem can be avoided since the contextual behavior of the senses is directly observable in the form of the local contexts of a word. From human disambiguation performance we know that the context of a word is usually sufficient to determine its sense. Based on this observation…