Dan Chazan

Technion Israel Institute of Technology, Electrical Engineering, Adjunct

Followers

Following

Co-authors

Public Views

Interests

Uploads

Papers by Dan Chazan

Efficient periodicity extraction based on sine-wave representation and its application to pitch determination of speech signals

This paper presents a novel low-complexity method for extracting periodicity of signals based on ... more This paper presents a novel low-complexity method for extracting periodicity of signals based on their sine-wave representation. In this representation, the signal is modeled as a finite sum of sine-waves, with time-varying amplitudes, phases and frequencies. We describe how one can modify the familiar spectral-comb analysis method to obtain a guaranteed and effective procedure to find the fundamental-frequency which gives the best harmonic approximation of the signal spectrum. The search is efficiently carried out in the frequency domain. The procedure obtains a successive refinement of possible pitch values which are consistent with an increasing number of sine wave components. Other pitch intervals are pruned at an early stage of the search. The advantage of this algorithm is its high accuracy achieved at a relatively low complexity. We also briefly describe one possible application in the area of pitch determination of speech signals.

Download

Building a speech database for the purpose of speaker specific speech synthesis

This paper presents practical and theoretical work carried out at IBM Research Laboratory, during... more This paper presents practical and theoretical work carried out at IBM Research Laboratory, during the course of a speech synthesis project. The paper deals with two separate issues. The first is the generation of a compact set of English utterances that will attain a good phonetic coverage of the language. The second issue is constructing a speaker specific database. This

Conversational networking: conversational protocols for transport, coding, and control

Improved Lower Bounds on Signal Parameter Estimation

IEEE Transactions on Information Theory, 1975

Download

Introduction to Approximation Theory (E. W. Cheney)

Siam Review, Jul 1, 1968

Download

Multiresolution time-frequency analysis of polyphonic music

ABSTRACT Automatic transcription of polyphonic music is one of the difficult problems in the fiel... more ABSTRACT Automatic transcription of polyphonic music is one of the difficult problems in the field of signal processing and analysis. One of the main reasons is that there exists no single scale of time-frequency representation, which is suitable for the detection of the wide range of features present in the musical sounds. The multiresolution Fourier transform (MFT) approach has been investigated, in order to overcome this problem by providing several time-frequency representations using a range of scales. This representation enables more flexibility in the analysis of polyphonic audio signals which requires a good feature separation. Transcription algorithms designed to work on MFT coefficients are shown to have a great potential and good transcription results are achieved for more complicated music performances, than currently possible by other techniques

Automatic transcription of polyphonic music using the multiresolution Fourier transform

The application of the multiresolution Fourier transform (MFT) has been investigated for the purp... more The application of the multiresolution Fourier transform (MFT) has been investigated for the purpose of analyses of polyphonic musical signals, and its potential for achieving reliable automatic transcription has been shown. This paper presents improved transcription algorithms which use the MFT coefficients for the recognition of a single polyphonic musical instrument. The algorithms have been tested on acoustic piano signals. Results show that reliable detection can be achieved if some limitations are imposed on the playing performance. Good results are attained in more complicated performances than those reported previously

Classification of transition sounds with application to automatic speech recognition

ABSTRACT This paper addresses the problem of classification of speech transition sounds. A number... more ABSTRACT This paper addresses the problem of classification of speech transition sounds. A number of non parametric classifiers are compared, and it is shown that some non-parametric classifiers have considerable advantages over traditional hidden Markov models. Among the non-parametric classifiers, support vector machines were found the most suitable and the easiest to tune. Some of the reasons for the superiority of non-parametric clas- sifiers will be discussed. The algorithm was tested on the voiced stop consonant phones extracted from the TIMIT corpus and re- sulted in very low error rates.

Method for encoding and decoding spectral phase data for speech signals

Journal of the Acoustical Society of America, 2007

A speech decoder and a segment aligner are provided in the present invention. The speech decoder ... more A speech decoder and a segment aligner are provided in the present invention. The speech decoder may include a spectrum reconstructor operative to reconstruct the spectrum of a speech segment from the amplitude envelope of the spectrum of said speech segment and pitch information, a phase combiner operative to reconstruct the complex spectrum of the speech segment from the reconstructed spectrum, phase information describing the speech segment, and pitch information describing the speech segment. The speech decoder may further include a delay operative to store a complex spectrum of a previous speech segment; and a segment aligner operative to determine the relative offset between the complex spectrum of the speech segment and the complex spectrum of the previous speech segment, align the position of the first pitch excitation of the current speech segment to the last pitch excitation of the previous speech segment; and to apply a time shift and a complex Hilbert filter to said complex spectra, wherein the segment aligner is operative to cross-correlate the complex spectra as C ⁡ ( τ ) = ∑ n = 0 N ⁢ ⁢ F n ⁢ G _ m ⁢ ⅇ - 2 ⁢ ⁢ π ⁢ ⁢ in ⁢ ⁢ τ , m = ⌊ n ⁢ p G p F + 0.5 ⌋ , where Fn and Gm are the computed complex magnitude of the pitch harmonics n and m of the current and previous spectra respectively, and pF and pG are their corresponding pitch periods.

Reducing the footprint of the IBM trainable speech synthesis system

This paper presents a novel approach for concatenative speech synthesis. This approach enables re... more This paper presents a novel approach for concatenative speech synthesis. This approach enables reduction of the dataset size of a concatenative text-to-speech system, namely the IBM trainable speech synthesis system, by more than an order of magnitude. A spectral acoustic feature based speech representation is used for computing a cost function during segment selection as well as for speech generation. Initial results indicate that even with a dataset size of a few megabytes it is possible to achieve quality which is significantly higher than existing small footprint formant based synthesizers.

Download

Automatic transcription of piano polyphonic music

ISPA .., 2005

ABSTRACT A novel algorithm for automatic transcription of piano polyphonic music is proposed. It ... more ABSTRACT A novel algorithm for automatic transcription of piano polyphonic music is proposed. It is based on a processing scheme that incorporates the following subtasks: segmentation of notes in time domain, estimation of frequency components based on the structure of time segments, extraction of pitches of underlying notes, and tracking of notes to obtain the final music score. A combination of multiresolution techniques, such as multiresolution Fourier transform and maximum likelihood frequency estimator, enables the user to successfully cope with the problems of constant time-frequency resolution and frequency masking. The algorithm demonstrates a better performance then results obtained by means of existing commercial software.

Eurospeech 2001- Scandinavia Plosive Spotting with Margin Classifiers

This paper presents a novel algorithm for precise spotting of plosives. The algorithm is based on... more This paper presents a novel algorithm for precise spotting of plosives. The algorithm is based on a pattern matching tech-nique implemented with margin classifiers, such as support vec-tor machines (SVM). A special hierarchical treatment to over-come the problem of fricative and false silence detection is pre-sented. It uses the loss-based multi-class decisions. Further-more, a method for smoothing the overall decisions by sequen-tial linear programming is described. The proposed algorithm was tested on the TIMIT corpus, which produced a very high spotting accuracy. The algorithm presented here is applied to plosives detection, but can easily be adapted to any class of phonemes. 1.

Discriminative kernel-based phoneme sequence recognition

We describe a new method for phoneme sequence recognition given a speech utterance, which is not ... more We describe a new method for phoneme sequence recognition given a speech utterance, which is not based on the HMM. In con-trast to HMM-based approaches, our method uses a discriminative kernel-based training procedure in which the learning process is tailored to the goal of minimizing the Levenshtein distance be-tween the predicted phoneme sequence and the correct sequence. The phoneme sequence predictor is devised by mapping the speech utterance along with a proposed phoneme sequence to a vector-space endowed with an inner-product that is realized by a Mercer kernel. Building on large margin techniques for predicting whole sequences, we are able to devise a learning algorithm which dis-tills to separating the correct phoneme sequence from all other sequences. We describe an iterative algorithm for learning the phoneme sequence recognizer and further describe an efficient im-plementation of it. We present initial encouraging experimental results with the TIMIT and compare the propos...

Download

Adaptive Homonic Noise Cancellation with Application of distribution Power lines Communications

4,506,380 3/1985 Matsui ................................. 381/71,04 5,018.202 5/1991 Takahashi et... more 4,506,380 3/1985 Matsui ................................. 381/71,04 5,018.202 5/1991 Takahashi et al. ................... 381/71,04 5,022,082 6/1991 Eriksson et al. ..................... 38/71.04 5.245,664 9/1993 Kinoshite et al. ...... ..., 38/71.04 5,251.262 I0/1993 Suzuki et al. ........................ 38A7104 5,384.853 1/1995 Kinoshita et al. ...... ... 381/71.04 5,410,604 4/1995 Saito et al. ........................... 38/7104 5,416,846 5/1995 Tamura et al. ....................... 38/04

Download

Discriminative kernel-based phoneme sequence recognition

We describe a new method for phoneme sequence recognition given a speech utterance, which is not ... more We describe a new method for phoneme sequence recognition given a speech utterance, which is not based on the HMM. In contrast to HMM-based approaches, our method uses a discriminative kernel-based training procedure in which the learning process is tailored to the goal of minimizing the Levenshtein distance between the predicted phoneme sequence and the correct sequence. The phoneme sequence predictor is devised by mapping the speech utterance along with a proposed phoneme sequence to a vector-space endowed with an inner-product that is realized by a Mercer kernel. Building on large margin techniques for predicting whole sequences, we are able to devise a learning algorithm which distills to separating the correct phoneme sequence from all other sequences. We describe an iterative algorithm for learning the phoneme sequence recognizer and further describe an efficient implementation of it. We present initial encouraging experimental results with the TIMIT and compare the proposed m...

Download

Phoneme alignment based on discriminative learning

We propose a new paradigm for aligning a phoneme sequence of a speech utterance with its acoustic... more We propose a new paradigm for aligning a phoneme sequence of a speech utterance with its acoustical signal counterpart. In contrast to common HMM-based approaches, our method employs a discriminative learning procedure in which the learning phase is tightly coupled with the alignment task at hand. The alignment function we devise is based on mapping the input acousticsymbolic representations of the speech utterance along with the target alignment into an abstract vector space. We suggest a specific mapping into the abstract vector-space which utilizes standard speech features (e.g. spectral distances) as well as confidence outputs of a framewise phoneme classifier. Building on techniques used for large margin methods for predicting whole sequences, our alignment function distills to a classifier in the abstract vector-space which separates correct alignments from incorrect ones. We describe a simple iterative algorithm for learning the alignment function and discuss its formal prope...

Download

Plosive spotting with margin classifiers

This paper presents a novel algorithm for precise spotting of plosives. The algorithm is based on... more This paper presents a novel algorithm for precise spotting of plosives. The algorithm is based on a pattern matching technique implemented with margin classifiers, such as support vector machines (SVM). A special hierarchical treatment to overcome the problem of fricative and false silence detection is presented. It uses the loss-based multi-class decisions. Furthermore, a method for smoothing the overall decisions by sequential linear programming is described. The proposed algorithm was tested on the TIMIT corpus, which produced a very high spotting accuracy. The algorithm presented here is applied to plosives detection, but can easily be adapted to any class of phonemes.

Download

Evaluation of adaptive multimicrophone algorithms for hearing aids

It appears that one of the most important factors limiting the usefulness of hearing aids is the ... more It appears that one of the most important factors limiting the usefulness of hearing aids is the high sensitivity of hearing aid users to interfering sounds. In this work, the possibility of utilizing computing power that could be packaged into a unit the size of a hearing aid in order to cancel out noises before they reach the ear was investigated. Algorithms for combining the outputs of a number of microphones so as to achieve a considerable noise reduction are proposed and evaluated, and some questions related to implementation are studied. Some experimental results based on numerical and listening tests are presented. The computation is started by indicating to the system a silent period of the main speaker. In order to minimize distortion for the main speaker, the algorithms try to produce an output that matches the first few autocorrelations of the main speaker. It is expected that the system should be able to track sufficiently slow variations in the positions of the noise so...

Download

Efficient periodicity extraction based on sine-wave representation and its application to pitch determination of speech signals

Download

A Large Margin Algorithm for Speech-to-Phoneme and Music-to-Score Alignment

Ieee Transactions on Audio Speech and Language Processing, Nov 1, 2007

We describe and analyze a discriminative algorithm for learning to align an audio signal with a g... more We describe and analyze a discriminative algorithm for learning to align an audio signal with a given sequence of events that tag the signal. We demonstrate the applicability of our method for the tasks of speech-to-phoneme alignment ("forced alignment") and music-to-score alignment. In the first alignment task, the events that tag the speech signal are phonemes while in the music alignment task, the events are musical notes. Our goal is to learn an alignment function whose input is an audio signal along with its accompanying event sequence and its output is a timing sequence representing the actual start time of each event in the audio signal. Generalizing the notion of separation with a margin used in support vector machines (SVM) for binary classification, we cast the learning task as the problem of finding a vector in an abstract inner-product space. To do so, we devise a mapping of the input signal and the event sequence along with any possible timing sequence into an abstract vector space. Each possible timing sequence therefore corresponds to an instance vector and the predicted timing sequence is the one whose projection onto the learned prediction vector is maximal. We set the prediction vector to be the solution of a minimization problem with a large set of constraints. Each constraint enforces a gap between the projection of the correct target timing sequence and the projection of an alternative, incorrect, timing sequence onto the vector. Though the number of constraints is very large, we describe a simple iterative algorithm for efficiently learning the vector and analyze the formal properties of the resulting learning algorithm. We report experimental results comparing the proposed algorithm to previous studies on speech-to-phoneme and music-to-score alignment, which use hidden Markov models (HMM). The results obtained in our experiments using the discriminative alignment algorithm are comparable to results of state-of-the-art systems.

Download

Dan Chazan

Uploads

Papers by Dan Chazan

Log In