Pitch Tracking

People identify powerfully with music: someone might say ÒthatÕs my song!Ó but they are unlikely to say ÒthatÕs my book!Ó or ÒthatÕs my picture!Ó A digital library of popular music therefore has the potential to be a compelling application of information retrieval technology. Such a library requires a retrieval method that is appropriate for a non-technical audience. Experiments on Òquery b y humming,Ó which attempt to retrieve a tune based on sampled recording of a user singing an excerpt, have heretofore concentrated on relatively small, well-curated collections. Scaling up introduces three problems: availability of source material, an increase in false positive hits, and slower retrieval. We describe our experiments with MIDI files, propose a new, more accurate distance metric between queries and songs, and discuss possibilities for efficient indexing.

Abstract-Chinese is known as a syllabic and tonal language and tone recognition plays an important role and provides very strong discriminative information for Chinese speech recognition [1]. Usually, the tone classification is based on the F0 (fundamental frequency) contours [2]. It is possible to infer a speaker's gender, age and emotion from his/her pitch, regardless of what is said. Meanwhile, the same sequence of words can convey very different meanings with variations in intonation. However, accurate pitch detection is difficult partly because tracked pitch contours are not ideal smooth curves. In this paper, we present a new smoothing algorithm for detected pitch contours. The effectiveness of the proposed method was shown using the HUB4-NE [3] natural speech corpus. Index Terms—Acoustic signal analysis, Acoustic signal detection I. INTRODUCTION OF PREVIOUS WORKS

This paper introduces a new spectral representation-based pitch estimation method. Since pitch is never stationary during real conversations, but often undergoes changes because of intonation, the spectral representation is derived from the Shorttime Harmonic Chirp Transform. This lets our technique to perform very well in noisy conditions, and to extract pitch values with high confidence, even from segments with strong intonations. The paper discusses a new way of segment-vice pitch extraction and does not deal with continuous pitch tracking, which is a topic of our future work. However, the performance of the proposed method is demonstrated on real recordings and the noise-dependency of its accuracy is numerically analyzed.

This paper proposes a new voicing detection and pitch estimation method that is particularly robust for noisy speech. This method is based on the spectral analysis of the speech multi-scale product. The multi-scale product (MP) consists of making the product of wavelet transform ...

This paper describes a peak-tracking spectrum analyzer, called Parshl, which is useful for extracting additive synthesis parameters from inharmonic sounds such as the piano. Parshl is based on the Short-Time Fourier Transform (STFT), adding features for tracking the amplitude, frequency, and phase trajectories of spectral lines from one FFT to the next. Parshl can be thought of as an “inharmonic phase vocoder” which uses tracking vocoder analysis channels instead of a fixed harmonic filter bank as used in previous FFT- ...

This paper describes a polyphonic note detection system incorporating a simple masking technique that can accurately transcribe chords and polyphonic piano music. The system, developed in MATLAB, will take input files in .wav format. The music is segmented by using Note Average Energy (NAE) onset detection. Onsets are used to segment the music into note windows which are then analysed using the FFT. Following compilation of the frequency peaks in each note window, an iterative masking procedure is used to detect and successively extract the notes. The masking procedure uses a database of note masks which are compiled from multiple note examples using both monophonic and polyphonic examples.

In this paper, a computationally efficient method for the estimation of the parameters of harmonic sinusoidal signals, including the order, which is of particular importance, for speech and audio signals is presented. The signal is modeled as a sum of harmonically related sinusoids in colored Gaussian noise. Aside from the order, the proposed method estimates the noise parameters, the fundamental frequency and the phase and amplitude of the individual harmonics. For the special case of white Gaussian noise, the method becomes particularly simple. The application of the proposed estimator to analysis of voiced speech and music signals is illustrated using real-life signals.

A pitch detection/tracking strategy for solo bowed-string and wind musical instrumental recordings is presented. To avoid the missing fundamental problem, we adopted the greatest common divisor method and modified it with a weighted-and-voting technique that can reveal more information of strong partials in the target signal. Moreover, a frame-based correction method with consideration of the performing aspects of the instruments is also proposed to emendate possible misjudgments in the transition from one note to the next note. Experimental results showed that the proposed strategy is superior to three popular methods for a pitch extraction/tracking task. The proposed method was also tested when the sound source is reverberant and the results were compared with other methods, too.

In this paper, the problem of legato pedalling technique detection in polyphonic piano music is addressed. We propose a novel detection method exploiting the effect of sympathetic resonance which can be enhanced by a legato-pedal onset. To measure the effect, a specific piano transcription was performed using the templates of pre-recorded isolated notes, from which partial frequencies were estimated. This promotes the acquisition of residual components associated to the weak coexcitation of damped notes due to the legato pedalling technique. Features that represent the sympathetic resonance measure were extracted from residuals. We finally used a logistic regression classifier to determine the existence of legato-pedal onsets.

Glottal instants namely GCIs and GOIs are useful in a wide variety of speech processing and biomedical applications. This paper presents the recent developments in the methodologies for glottal activity detection using EGG and speech signals. The databases used by the researchers are APLAWD, SAM, KEELE and BDL. The methodologies are compared in terms of identification rate, miss rate, and false alarm rate and identification accuracy. The applications in which these methods are employed are also presented. It is observed that many researches are conducted to identify the glottal activity based on the speech signal rather than the EGG signal.

We prove that a finite (state and action spaces) semi-Markov decision process with limiting ratio average (undiscounted) payoff has an optimal pure semi-stationary policy (i.e., a semi-Markov policy independent of decision epoch count). We conclude by showing (with the aid of an example) that the result cannot be strengthened further. A crude but finite step algorithm is given to compute such an optimal policy.

In this paper, the source-filter model of speech production is adapted to represent the speech signal as the superposition and convolution of a dynamic source and resonant modes. The aim is to increase the resolution of the time-instantaneous-frequency representation of each of the individual contributions of different sections of the human phonatory system. We present a framework based on dynamic mode predictors and filters, which are adapted, using gradient-based techniques, to track the modal dynamics of speech yielding a representation which is free from quasi-stationary assumptions thus allowing flexible manipulation of the speech signal. Several examples are offered including intonation modifications to illustrate the potential of the proposed approach.

In this paper the improvement in performance of automatic speech recognition (ASR) system is achieved with help of pitch dependent features and probability of voicing estimated features. The pitch dependent features are useful for tonal language ASR system. Punjabi language is highly tonal language and hence here we are building ASR system for Punjabi language with pitch dependent features and probability of voicing estimated features. The word error rate of system gives the performance of system which drastically improves with pitch dependent features and probability of voicing estimated features. Comparison of Yin, SAcC, Fundamental Frequency Variation (FFV) and Kaldi pitch features of ASR system were done in terms of WER. The KALDI pitch tracker of Kaldi toolkit gives the best performance ASR system among other featured ASR systems. The performance of ASR system is evaluated for Punjabi language.

We propose in this paper a spectral synthesis model to generate noisy sounds with independent control parameters for spectral density and spectral envelope. Algorithms defining in a efficient way these spectral properties from the statistical parameters of the model are presented in details. Real-time implementation is composed of many methods which can be sequenced.

This paper describes a polyphonic note detection system incorporating a simple masking technique that can accurately transcribe chords and polyphonic piano music. The system, developed in MATLAB, will take input files in .wav format. The music is segmented by using Note Average Energy (NAE) onset detection. Onsets are used to segment the music into note windows which are then analysed using the FFT. Following compilation of the frequency peaks in each note window, an iterative masking procedure is used to detect and successively extract the notes. The masking procedure uses a database of note masks which are compiled from multiple note examples using both monophonic and polyphonic examples.

In this thesis a novel multiresolution approach for note detection in a polyphonic mix is proposed. The idea is to use a set of wavelets whose lengths are adapted to the theoretical fundamental period of musical notes. Using the typical wavelet dyadic decomposition we can generate a set of wavelets that match the fundamental frequency (F0) of a given note in every octave. Therefore, using a set of 12 different wavelets, one per each semitone, we can represent the fundamental frequency of every note in every octave using one wavelet scale per each octave. The magnitude and phase continuity of wavelet coefficients across temporal frames is exploited to draw a special kind of spectrogram, namely Pitch-Synchronous Wavelet Spectrogram (PSWS). When the corresponding F0 and harmonics of a note are present in the signal, a special DC pattern appears in the PSWS, due to the aforementioned continuity. Any other harmonic signal or noise produces pseudo-periodic or random AC patterns. This way,...

This paper describes a sound synthesis technique that modulates the coefficients of allpass filter chains using audio-rate frequencies. It was found that modulating a single allpass filter section produces a feedback AM–like spectrum, and that its bandwidth is extended and further processed by non-sinusoidal FM when the sections are cascaded. The cascade length parameter provides dynamic bandwidth control to prevent upper range aliasing artifacts, and the amount of spectral content within that band can be controlled ...

In this paper, we describe the BBN 2007 Mandarin Speechto-Text system developed for the GALE Evaluation 2007. In comparison to the BBN 2006 Mandarin system, we achieved 25% relative reduction in character error rate on the most important test sets. The utilization of all available training data provided the largest contribution to the improvement. The use of a better pitch tracking algorithm also contributed sig-nicantly, while system combination made some noticeable improvement too.

Music understanding is a process closely related to the knowledge and experience of the listener. The amount of knowledge required is relative to the complexity of the task in hand. This dissertation is concerned with the problem of automatically decomposing musical signals into a score-like representation. It proposes that, as with humans, an automatic system requires knowledge about the signal and its expected behaviour to correctly analyse music. The proposed system uses the blackboard architecture to combine the use of knowledge with data provided by the bottom-up processing of the signal's information. Methods are proposed for the estimation of pitches, onset times and durations of notes in simple polyphonic music. A method for onset detection is presented. It provides an alternative to conventional energy-based algorithms by using phase information. Statistical analysis is used to create a detection function that evaluates the expected behaviour of the signal regarding onsets. Two methods for multi-pitch estimation are introduced. The first concentrates on the grouping of harmonic information in the frequency-domain. Its performance and limitations emphasise the case for the use of high-level knowledge. This knowledge, in the form of the individual waveforms of a single instrument, is used in the second proposed approach. The method is based on a time-domain linear additive model and it presents an alternative to common frequency-domain approaches. Results are presented and discussed for all methods, showing that, if reliably generated, the use of knowledge can significantly improve the quality of the analysis. A mi casa, la que no tiene ni paredes ni techo, sino a Salvador, Maritza, Jesús y la Mava. Completing my doctoral research in a foreign country and in a language which is not my own has been quite an experience, one that can be moderately described as "intense". This dissertation has only been possible given the immense support and cooperation that I received from family, friends and colleagues. For them, to whom I am so profoundly indebted, I dedicate the following remarks. First of all, I would like to thank my supervisor Prof. Mark Sandler for opening the doors to a life-changing experience. For having the courage of recruiting me when none of my words made any sense. For the continuous practical and emotional support. For the good advise during the length of this project. Thanks to the Joint Information Systems Committee (JISC) in the United Kingdom, the National Science Foundation (NSF) in the United States and the Fundación Gran Mariscal de Ayacucho in Venezuela for funding my research and for covering my living expenses in London during the last three years.

In this work, a new instantaneous fundamental frequency extraction method is presented, with the attention especially focused on its robustness for pathological voices processing. It is based on the Ensemble Empirical Mode Decomposition (EEMD) algorithm, which is a completely datadriven method for signal decomposition into a sum of AM-FM components, called Intrinsic Mode Functions (IMFs) or modes. Our results show that the speech fundamental frequency can be captured in a single IMF. We also propose an algorithm for selecting the mode where the fundamental frequency can be found, based on the logarithm of the power of the IMFs. The instantaneous frequency is then extracted by means of well-known techniques. The behaviour of the proposed method is compared with other two ones (Robust Algorithm for Pitch Tracking-RAPT-and auto-correlation based algorithms), both in normal and pathological sustained vowels.

This thesis examines applications of Digital Signal Processing to the analysis, transformation, and resynthesis of musical audio. First I give an overview of the human perception of music. I then examine in detail the requirements for a system that can analyse, transcribe, process, and resynthesise monaural polyphonic music. I then describe and compare the possible hardware and software platforms. After this I describe a prototype hybrid system that attempts to carry out these tasks using a method based on .. additive synthesis. Next I present results from its application to a variety of musical examples, and critically assess its pert:ormance and limitations. I then address these issues in the design of a second system based on Gabor wavelets. I conclude by summarising the research and outlining suggestions for future developments.

In this paper, we describe the BBN 2007 Mandarin Speechto-Text system developed for the GALE Evaluation 2007. In comparison to the BBN 2006 Mandarin system, we achieved 25% relative reduction in character error rate on the most important test sets. The utilization of all available training data provided the largest contribution to the improvement. The use of a better pitch tracking algorithm also contributed sig-nicantly, while system combination made some noticeable improvement too.

Close to glottal closure instants (GCIs), the speech signal is expected to change its amplitude rapidly and, at GCIs, it is expected to have strong negative peaks. A novel algorithm that exploits these two properties for the estimation of GCIs is presented. Here, a symmetrised speech segment is assumed to be a Fourier transform (FT) of an even function. In such a case, at the locations of the GCIs, the strong negative peaks in the symmetrised speech segment correspond to zeros that lie considerably outside the unit circle in the z-plane. The group delay spectrum of the time-domain signal derived by taking inverse FT of this assumed FT is expected to take a value close to −2π at the angular locations of these zeros. Mapping frequency scale to time scale, the frequency bins for which group delay reaches −2π correspond to the locations of GCIs. Theoretical justification for the proposed approach is also presented by defining a novel function called the conditional group delay function. Systematic evaluation is carried out on the CMU Arctic database and the performance of the proposed technique is better than that of the algorithms namely DYPSA, ZFF, YAGA and is close to that of SEDREAMS.

A new light is thrown on the Portnoff [1] speech signal timescale modification algorithm. It is shown in particular that the Portnoff algorithm easily accommodates expansion factors bigger than 2 without causing reverberation nor chorusing. The modified Portnoff algorithm, which draws on spectral modification techniques due to Seneff [2], has been tested on several speech signals. The quality of the synthesized signal is totally satisfactory even for big expansion factors. The article gives a brief summary of the Portnoff algorithm and spells out the modifications introduced. It is shown that the phase unwrapping procedure constitutes a crucial point of the algorithm.

The aim of the paper is to show a system engineered for automatic detection and correction of detuned singing. For this purpose, existing methods of fundamental frequency detection and pitch correction are reviewed. In addition, main characteristics of some existing detuning systems are presented. As algorithms for fundamental frequencies detection and pitch correction, the fast autocorrelation and HPS (Harmonic Product Spectrum), and the modified phase vocoder and PSOLA (Pitch-Synchronous Overlap-Add) are chosen and examined. Four possible combinations of the algorithms are reviewed not only in the context of fundamental frequency detection and pitch shifting correctness but also with regard to the quality of the resulting singing signal. Experiments are performed on both male and female singing samples consisting of a variety of tones and various articulations. Basing on the obtained results, it is concluded that the HPS and PSOLA algorithms are the optimum choice as means to corr...

Although research on the use and effectiveness of visual feedback for teaching tone and intonation began more than thirty years ago, the technology for signal analysis and pitch extraction using microcomputers has only recently become widely accessible and affordable. This paper 1) reviews the major pedagogical applications of acoustic phonetic research for teaching segmentals (individual sounds) and suprasegmentals (intonation, stress, rhythm); 2) summaries the hardware and software currently available for speech analysis on the Macintosh and IBM-PCs, and 3) discusses courseware features that should be included in implementing this new technology to help foreign language students improve their production and perception of tone and intonation.

In this paper, the source-filter model of speech production is adapted to represent the speech signal as the superposition and convolution of a dynamic source and resonant modes. The aim is to increase the resolution of the time-instantaneous-frequency representation of each of the individual contributions of different sections of the human phonatory system. We present a framework based on dynamic mode predictors and filters, which are adapted, using gradient-based techniques, to track the modal dynamics of speech yielding a representation which is free from quasi-stationary assumptions thus allowing flexible manipulation of the speech signal. Several examples are offered including intonation modifications to illustrate the potential of the proposed approach.

In this paper, we introduce a pitch detection algorithm that is particularly robust for telephone speech and prosodic modeling. The algorithm uses a logarithmically sampled spectral representation of speech, similar to that in the subharmonic summation approach. Constraints for logF0 and ΔlogF0 are combined in a dynamic programming search to find an optimum pitch track. The search algorithm is able to find a continuous pitch contour regardless of the voicing status, while a separate voicing decision module computes the probability of voicing per frame. We evaluated the algorithm using the Keele pitch extraction reference database under both studio and telephone conditions. Our algorithm is very robust to channel degradation, and compares favorably to XWAVES under telephone conditions. It also significantly outperforms XWAVES when used for tone classification on a telephone quality Mandarin digit corpus

In this work, a new instantaneous fundamental frequency extraction method is presented, with the attention especially focused on its robustness for pathological voices processing. It is based on the Ensemble Empirical Mode Decomposition (EEMD) algorithm, which is a completely data-driven method for signal decomposition into a sum of AM - FM components, called Intrinsic Mode Functions (IMFs) or modes. Our results show that the speech fundamental frequency can be captured in a single IMF. We also propose an algorithm for selecting the mode where the fundamental frequency can be found, based on the logarithm of the power of the IMFs. The instantaneous frequency is then extracted by means of well-known techniques. The behaviour of the proposed method is compared with other two ones (Robust Algorithm for Pitch Tracking -RAPT- and auto-correlation based algorithms), both in normal and pathological sustained vowels.

The problem of music retrieval by sung query consists of building a machine capable of simulating the cognitive process of identifying a musical piece from a few sung notes of its melody. In this paper, the algorithms of pitch tracking, onset detection and melody matching used in the system Tararira are described. Much effort has been put on automatic transcription of singing voice as it is a key factor in the overall performance. A novel way of combining note by note matching with a recent approach based on pitch time series matching is introduced. Resumen. El problema de búsqueda de música por tarareo consiste en construir un sistema capaz de simular el proceso cognitivo de identificar una pieza musical a partir de unas pocas notas cantadas de su melodía. En este artículo se describen los algoritmos de detección de altura, segmentación de audio en notas y comparación de melodías utilizados en el sistema Tararira. Se concentran esfuerzos en la transcripción automática de la voz cantada ya que es determinante en el desempeño del sistema. Para la comparación de melodías se propone una forma de combinar los enfoques basados en notas y series temporales, considerados antagónicos hasta el momento.

This extended abstract details a submission to the Music In-formation Retrieval Evaluation eXchange in the Query by Singing/Humming task. The problem of query by singing consists of building a machine capable of simulating the cognitive process of identifying a musical piece from a few sung notes of its melody. In this work, the algorithms of pitch tracking, onset detection and melody matching used in the system Tararira [1] are briefly described. Much effort has been put on automatic transcription of singing voice as it is a key factor in the overall performance. A novel way of combining note by note matching with the approach based on pitch time series matching is introduced.

An effective multi-pitch tracking algorithm for noisy speech is critical for auditory processing. However, the performance of existing algorithms is not satisfactory. We have developed a robust algorithm for multi-pitch tracking of noisy speech based on statistical anticipation. By combining an improved channel and peak selection method, a new integration method for extracting periodicity information across the different channels, and a hidden Markov model (HMM) for forming continuous pitch tracks, our algorithm can reliably track single and double pitch tracks in a noisy environment.

We present a robust algorithm for multi-pitch tracking of noisy speech. Our approach integrates an improved channel and peak selection method, a new integration method for extracting periodicity information across different frequency channels, and a hidden Markov model (HMM) for forming continuous pitch tracks, and as a result, our algorithm can reliably track single and double pitch tracks in a noisy environment. The proposed algorithm is evaluated on a database of speech utterances mixed with various interferences and the results show that our algorithm outperforms existing algorithms significantly.

Neural processes underlying pitch perception at the level of the cerebral cortex are influenced by language experience. We investigated whether early, pre-attentive stages of pitch processing at the level of the human brainstem may also be influenced by language experience. The human frequency following response (FFR), reflecting sustained phase-locked activity in a population of neural elements, was used to measure activity within the rostral brainstem. FFRs elicited by four Mandarin tones were recorded from native speakers of Mandarin Chinese and English. Pitch strength (reflecting robustness of neural phase-locking at the pitch periods) and accuracy of pitch tracking were extracted from the FFRs using autocorrelation algorithms. These measures revealed that the Chinese group exhibits stronger pitch representation and smoother pitch tracking than the English group. Consistent with the pitch data, FFR spectral data showed that the Chinese group exhibits stronger representation of t...

Neural encoding of pitch in the auditory brainstem is shaped by long-term experience with language. The aim herein was to determine to what extent this experience-dependent effect is specific to a particular language. Analysis of variance of brainstem responses to Mandarin and Thai tones revealed that regardless of language identity, pitch-tracking accuracy of whole tones was higher in the two tone language groups (Chinese, Thai) compared to the non-tone language group (English), and that pitch strength of 40-ms tonal sections was generally more robust in tone relative to non-tone languages. Discriminant analysis of tonal sections, as defined by variation in direction and degree of slope, showed that moderate rising pitch was the most important variable for classifying English, Chinese, and Thai participants into their respective groups. We conclude that language-dependent enhancement of pitch representation transfers to other languages with similar phonological systems. From a neurobiological perspective, these findings suggest that neural mechanisms local to the brainstem are tuned for processing pitch dimensions that are perceptually salient depending upon the melodic patterns of a language.

This thesis examines applications of Digital Signal Processing to the analysis, transformation, and resynthesis of musical audio. First I give an overview of the human perception of music. I then examine in detail the requirements for a system that can analyse, transcribe, process, and resynthesise monaural polyphonic music. I then describe and compare the possible hardware and software platforms. After this I describe a prototype hybrid system that attempts to carry out these tasks using a method based on .. additive synthesis. Next I present results from its application to a variety of musical examples, and critically assess its pert:ormance and limitations. I then address these issues in the design of a second system based on Gabor wavelets. I conclude by summarising the research and outlining suggestions for future developments.

This paper introduces the Vector Phaseshaping (VPS) synthesis technique, which extends the classic Phase Distortion method by providing flexible means to distort the phase of a sinusoidal oscillator. This is achieved by describing the phase distortion function using one or more breakpoint vectors, which are then manipulated in two dimensions to produce waveshape modulation at control and audio rates. The synthesis parameters and their effects are explained, and the spectral description of the method is derived. Certain synthesis parameter combinations result in audible aliasing, which can be reduced with a novel aliasing suppression algorithm described in the paper. The extension is capable of producing a variety of interesting harmonic and inharmonic spectra, including for instance, formant peaks, while the two-dimensional form of the control parameters is expressive and is well suited for interactive applications.

We prove that a finite (state and action spaces) semi-Markov decision process with limiting ratio average (undiscounted) payoff has an optimal pure semi-stationary policy (i.e., a semi-Markov policy independent of decision epoch count). We conclude by showing (with the aid of an example) that the result cannot be strengthened further. A crude but finite step algorithm is given to compute such an optimal policy.

We propose a method for segmentation of pitch tracks for melody detection in polyphonic musical signals. This is an important issue for melody-based music information retrieval, as well as melody transcription. Past work in the field addressed especially the issue of extracting melodic pitch lines, without explicit definition of notes. Thus, in this work, we propose a two-stage segmentation of pitch tracks for the determination of musical notes. In the first stage, frequency-based segmentation is conducted with recourse to frequency variations in pitch tracks. In the second phase, salience-based segmentation is performed in order to split consecutive notes with equal value, using pitch salience minima and note onsets.

In this paper, the source-filter model of speech production is adapted to represent the speech signal as the superposition and convolution of a dynamic source and resonant modes. The aim is to increase the resolution of the time-instantaneous-frequency representation of each of the individual contributions of different sections of the human phonatory system. We present a framework based on dynamic mode predictors and filters, which are adapted, using gradient-based techniques, to track the modal dynamics of speech yielding a representation which is free from quasi-stationary assumptions thus allowing flexible manipulation of the speech signal. Several examples are offered including intonation modifications to illustrate the potential of the proposed approach.

In this paper, we are proposing the idea of making an automated software that will transcribe each note while the musician plays the instrument. The software will take the sound of the instrument as an input and will process the frequency of each note and the way it is played and transcribe the note visually. In this project, we will consider recording consisting of only monophonic notes i.e. only a single note will be played at a time.This paper focuses on extracting audio, detecting pitch and displaying symbols. The project makes extent use of audio signal processing and python libraries (pyAudio).. IndexTerms audio signal processing, pyAudio.

El fallecimiento de Eduardo Lizalde (1929-2022) representa una pérdida irreemplazable para la poesía mexicana. A lo largo de su trayectoria literaria, desarrolló una afinidad casi inevitable con la sonoridad de su voz, ya fuera a través de sus lecturas poéticas o al discutir la ópera que tanto admiraba en sus cápsulas de Opus 94.5 FM y otras estaciones radiofónicas de la Ciudad de México. Para fortuna del patrimonio histórico y literario del país, existen numerosas grabaciones audiovisuales que permitirán a futuras generaciones experimentar con sus propios sentidos lo que significaba verlo y escucharlo recitar de viva voz. En PoéticaSonora MX le ofrecemos este homenaje póstumo desde la crítica literaria con un enfoque en la auralidad de la palabra entonada en un contexto creativo-literario.

Background Children with pervasive developmental disorders (PDD), such as children with autism spectrum disorders (ASD), often show auditory processing deficits related to their overarching language impairment. Auditory training programs such as Fast ForWord Language may potentially alleviate these deficits through training-induced improvements in auditory processing. Methods To assess the impact of auditory training on auditory function in children with ASD, brainstem and cortical responses to speech sounds presented in quiet and noise were collected from five children with ASD who completed Fast ForWord training. Results Relative to six control children with ASD who did not complete Fast ForWord, training-related changes were found in brainstem response timing (three children) and pitch-tracking (one child), and cortical response timing (all five children) after Fast ForWord use. Conclusions These results provide an objective indication of the benefit of training on auditory funct...

In this paper, we describe the BBN 2007 Mandarin Speechto-Text system developed for the GALE Evaluation 2007. In comparison to the BBN 2006 Mandarin system, we achieved 25% relative reduction in character error rate on the most important test sets. The utilization of all available training data provided the largest contribution to the improvement. The use of a better pitch tracking algorithm also contributed sig-nicantly, while system combination made some noticeable improvement too.

We propose a method for segmentation of pitch tracks for melody detection in polyphonic musical signals. This is an important issue for melody-based music information retrieval, as well as melody transcription. Past work in the field addressed especially the issue of extracting melodic pitch lines, without explicit definition of notes. Thus, in this work, we propose a two-stage segmentation of pitch tracks for the determination of musical notes. In the first stage, frequency-based segmentation is conducted with recourse to frequency variations in pitch tracks. In the second phase, salience-based segmentation is performed in order to split consecutive notes with equal value, using pitch salience minima and note onsets.

and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, redistribution , reselling , loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

This paper presents a voiced/unvoiced classification algorithm of the noisy speech signal by analyzing two acoustic features of the speech signal. Short-time energy and short-time zero-crossing rates are one of the most distinguishable time domain features of a speech signal to classify its voiced activity into voiced/unvoiced segment. A new idea is developed where frame by frame processing has done in narrow band speech signal using spectrogram image. Two time domain features, short-time energy (STE) and short-time zero-crossing rate (ZCR) are used to classify its voiced/unvoiced parts. In the first stage, each frame of the analyzing spectrogram is divided into three separate sub bands and examines their short-time energy ratio pattern. Then an energy ratio pattern matching look up table is used to classify the voicing activity. However, this method successfully classifies patterns 1 through 4 but fails in the rest of the patterns in the look up table. Therefore, the rest of the patterns are confirmed in the second stage where frame wise short-time average zero-crossing rate is compared with a threshold value. In this study, the threshold value is calculated from the short-time average zero-crossing rate of White Gaussian Noise (wGn). The accuracy of the proposed method is evaluated using both male and female speech waveforms under different signal-to-noise ratios (SNRs). Experimental results show that the proposed method achieves better accuracy than the conventional methods in the literature.