Browsing by Author "Chang, WW"

Now showing 1 - 20 of 20

A hybrid strategy for Gilbert's channel characterization using gradient and annealing techniques
(1998-06-01)
Tan, TH
;
Chang, WW
;
電信工程研究所
;
Institute of Communications Engineering
This paper presents a hybrid algorithm for estimating Gilbert's channel model parameters from an experimental error-gap distribution. A stochastic simulated annealing algorithm is applied to determine automatically a set of good starting points, which are then used by the deterministic gradient algorithm for faster convergence to the global optimum. Simulation results indicate that, for channel characterization, this hybrid strategy provides an ideal compromise between modelling accuracy and convergence time.
A masking-threshold-adapted weighting filter for excitation search
(1996-03-01)
Chang, WW
;
Wang, CT
;
交大名義發表
;
電信工程研究所
;
National Chiao Tung University
Most LPC-based audio coders improve reproduction quality by using predictor coefficients to embody perceptual masking in noise spectral shaping. Since the predictor coefficients were originally derived to characterize sound production models, they cannot precisely describe the human ear's nonlinear responses to frequency and loudness. In this paper, we report on new approaches to exploiting the masking threshold in the design of a perceptual noise-weighting filter for excitation searches. To track the nonstationary evolution of a masking threshold, an autoregressive spectral analysis with finite order has been shown to be capable of providing sufficient accuracy. In seeking faster response, an artificial neural network was also trained to extract autoregressive modeling parameters of the masking threshold from typical audio signals via mapping. Furthermore, we propose the concept of sinusoidal excitation representation to better track the intrinsic characteristics of prediction error signals. Simulation results indicate that the combined use of a multisinusoid excitation model and a masking-threshold-adapted weighting filter allows the implementation of an LPC-based audio coder that delivers near transparent quality at the rate of 96 kb/s.
Audio coding using sinusoidal excitation representation
(1997)
Chang, WW
;
Wang, DY
;
Wang, LW
;
交大名義發表
;
電信工程研究所
Most LPC-based audio coders employ simplistic noise-shaping operations to perform psychoacoustic control of quantization noise. In this paper, we report on new approaches to exploiting perceptual masking in the design of adaptive quantization of LPC excitation parameters. Due to its localized spectral sensitivity, sinusoidal excitation representation is preferred to spectrally flat signals for use in excitation modeling. Simulation results indicate that the proposed multisinusoid excited coder can deliver high quality audio reproduction at the rate of 72 kb/s.
Chinese dialect identification based on genetic algorithm for discriminative training of bigram model
(2000-12-01)
Tsai, WH
;
Chang, WW
;
電信工程研究所
;
Institute of Communications Engineering
A minimum classification error formulation based on genetic algorithm is proposed for discriminative training of the bigram language model. Results of Chinese dialect identification were reported which demonstrate performance improvement with use of the genetic algorithm over the generalized probabilistic descent algorithm.
Chinese dialect identification using segmental and prosodic features
(2000-10-01)
Chang, WW
;
Tsai, WH
;
電信工程研究所
;
Institute of Communications Engineering
Several approaches to Chinese dialect identification based on segmental and prosodic features of speech are described in this paper. When using segmental information only, the system performs phonotactic analysis after speech utterances have been tokenized into sequences of broad phonetic classes. The second scheme comprises prosodic models which are trained to capture tone sequence information for individual dialects. Also proposed is a novel approach that examines differences between Chinese dialects at broad phonetic and prosodic levels. These algorithms were evaluated via a multispeaker read-speech mode. Simulation results indicate that the combined use of Segmental and prosodic features allows the proposed system to discriminate among three major Chinese dialects spoken in Taiwan with 93.0% accuracy. (C) 2000 Acoustical Society of America. [S0001-4966(oo)03410-X].
Constrained VQ codebook design for noisy channels
(2001-05-10)
Chang, WW
;
Hsu, HI
;
電信工程研究所
;
Institute of Communications Engineering
A codebook design approach for constrained vector quantisation using the Hadamard transform of channel transition probabilities is proposed. It is examined for quantisation of Gauss-Markov sources over channels with memory and compared with the generalised Lloyd algorithm.
Discriminative training of Gaussian mixture bigram models with application to Chinese dialect identification
(2002-03-01)
Tsai, WH
;
Chang, WW
;
電信工程研究所
;
Institute of Communications Engineering
This study focuses on the parametric stochastic modeling of characteristic sound features that distinguish languages from one another. A new stochastic model. the so-called Gaussian mixture bigram model (GMBM), that allows exploitation of the acoustic feature bigram statistics without requiring transcribed training data is introduced. For greater efficiency, a minimum classification error (MCE) algorithm is employed to accomplish discriminative training of a GMBM-based Chinese dialect identification system. Simulation results demonstrate the effectiveness of the GMBM for dialect-specific acoustic modeling, and use of this model allows the proposed system to distinguish between the three major Chinese dialects spoken in Taiwan with 94.4% accuracy. (C) 2002 Elsevier Science B.V. All rights reserved.
Estimation of Gilbert's model parameters using the simulated annealing method
(1996)
Tan, TH
;
Chang, WW
;
交大名義發表
;
電信工程研究所
;
National Chiao Tung University
Errors encountered in digital mobile radio channels are not independent but exhibit varying degrees of statistical dependencies. In many mobile applications, Gilbert's finite-state Markov chain model has been shown to adequately characterize the bursty-noise binary channels. An estimation method based on the simulated annealing algorithm is proposed for computing the model parameters from the sample error sequences. Experimental results indicate that the simulated annealing method yields more accurate identification at a faster convergence rate for comparison with the gradient iterative method.
Estimation of the Gilbert model parameters using the simulated annealing method
(1996-07-04)
Tan, TH
;
Chang, WW
;
交大名義發表
;
電信工程研究所
;
National Chiao Tung University
An estimation method based on the simulated annealing algorithm is proposed for computing the Gilbert model parameters from sample error sequences. Experimental results indicate that the simulated annealing method yields estimates which more closely match the experimental error statistics.
Index assignment in vector quantisation for channels with memory
(2002-06-01)
Chang, WW
;
Hsu, HI
;
電信工程研究所
;
Institute of Communications Engineering
The study presents a formulation of a Hadamard framework for analysing the vector quantisation over channels with memory. In seeking faster response, classes of index assignments are defined in terms of the Hadamard transform of channel transition probabilities. An index assignment algorithm is developed that achieves high robustness against channel errors, and its performance in vector quantisation of Gauss-Markov sources under noisy channel conditions is illustrated.
Memory-enhanced MMSE decoding in vector quantization
(2003-10-01)
Hsu, HL
;
Chang, WW
;
Liu, XB
;
Koh, SN
;
電信工程研究所
An approach to minimum mean-squared error (MMSE) decoding for vector quantization over channels with memory is presented. The decoder is based on the Gilbert channel model that allows the exploitation of both intra- and inter-block correlation of bit error sequences. We also develop a recursive algorithm for computing the a posteriori probability of a transmitted index sequence, and illustrate its performance in quantization of Gauss-Markov sources under noisy channel conditions.
MMSE decoding for vector quantization over channels with memory
(2002)
Hsu, HI
;
Chang, WW
;
Liu, XB
;
Koh, SN
;
電信工程研究所
This paper presents memory-enhanced extensions of minimum mean-squared error (MMSE) decoding for vector quantization over noisy channels. We also develop a recursive algorithm for computing the transition probabilities of the Gilbert channel, and illustrate its performance in vector quantization of Gauss-Markov sources under noisy channel conditions. Simulation results indicate that the proposed algorithm enables the implementation of an MMSE decoder with increased robustness to channel errors.
Perceptual quantisation of LPC excitation parameters
(1998-06-01)
Chang, WW
;
Wang, DY
;
電信工程研究所
;
Institute of Communications Engineering
A new approach to wideband digital audio compression at 64 kbit/s is presented. This LPC coder features a sinusoidal excitation model and adaptive bit-allocation based on a mask-to-noise ratio perceptual measure. Owing to its localised spectral sensitivity, sinusoidal excitation representation can be shown to provide an ideal framework for incorporating an estimated masking threshold in the design of noise spectral shaping. Quantisation of LPC and excitation parameters is also examined. Simulation results indicate that sine waves are preferred to spectrally flat signals for use in excitation modelling, because in the former the estimated masking threshold can be more precisely implemented in the distribution of quantisation noise.
Quality enhancement of sinusoidal transform vocoders
(1998-12-01)
Chang, WW
;
Wang, DY
;
電信工程研究所
;
Institute of Communications Engineering
The authors present quality enhancement of sinusoidal transform coders (STCs) via the development of parametric models. The benefits of the Bark spectrum are explored for use in the design of perceptual coding of the sine-wave amplitudes. According to the results, the proposed approach provides a uniform perceptual fit across the spectrum. To enhance the accuracy of phase representation, noncausal all-pole modelling of the vocal system is also discussed. Experimental results indicate that the use of the developed new parametric models allows the STC to improve the phase accuracy as well as the synthetic speech quality.
Robust vector quantization for wireless channels
(2001-07-01)
Chang, WW
;
Tan, TH
;
Wang, DY
;
電信工程研究所
;
Institute of Communications Engineering
This study focuses on two issues: parametric modeling of the channel and index assignment of codevectors, to design a vector quantizer that achieves high robustness against channel errors. We first formulate the design of a robust zero-redundancy vector quantizer as a combinatorial optimization problem leading to a genetic search for a minimum-distortion index assignment. Performance is further enhanced by the use of the Fritchman channel model that more closely characterizes the statistical dependencies between error sequences. This study also presents an index assignment algorithm based on the Fritchman model with parameter values estimated using a real-coded genetic algorithm. Simulation results indicate that the global explorative properties of genetic algorithms make them very effective in estimating Fritchman model parameters, and use of this model can match index assignment to expected channel conditions.
Spectral and prosodic transformations of hearing-impaired Mandarin speech
(2006-02-01)
Lee, CL
;
Chang, WW
;
Chiang, YC
;
電信工程研究所
;
Institute of Communications Engineering
This paper studies the combined use of spectral and prosodic conversions to enhance the hearing-impaired Mandarin speech. The analysis-synthesis system is based on a sinusoidal representation of the speech production mechanism. By taking advantage of the tone structure in Mandarin speech, pitch contours are orthogonally transformed and applied within the sinusoidal framework to perform pitch modification. Also proposed is a time-scale modification algorithm that finds accurate alignments between hearing-impaired and normal utterances. Using the alignments, spectral conversion is performed on subsyllabic acoustic units by a continuous probabilistic transform based on a Gaussian mixture model. Results of perceptual evaluation indicate that the proposed system greatly improves the intelligibility and the naturalness of hearing-impaired Mandarin speech. (c) 2005 Elsevier B.V. All rights reserved.
Statistical characterization of error sequences and its applications to error control
(1996)
Chang, WW
;
Tan, TH
;
交大名義發表
;
電信工程研究所
;
National Chiao Tung University
Average bit error rate is not a satisfactory descriptive feature required for code evaluation, because it can not precisely distinguish between random and bursty natures inherent in error occurrences. In this paper we report new results of exploiting the statistical characterization of error sequences in the design of error control for mobile radio communications. For tracking the nonstationary evolution of error sequences, Gilbert's model is considered here as a model representative of real channel behavior. We first employ the simulated annealing technique to extract the statistical parameters that are required for descriptive modeling the error sequences. To exploit this through adaptive coding system, we also construct a class of cyclic codes by using the least common multiple of the generator polynomials of both the BCH and Five codes.
Techniques for improving sinusoidal transform vocoders
(1998)
Chang, WW
;
Wang, DY
;
電信工程研究所
;
Institute of Communications Engineering
This paper presents quality enhancement of sinusoidal transform coders (STC) via the development of new parametric models. First explored are the benefits of Bark spectrum for use in the design of perceptual coding of the sine-wave amplitudes. According to our results, the proposed approach provides a uniform perceptual fit across the spectrum. To enhance the accuracy of phase representation, noncausal all-pole modeling of the vocal system is also discussed. Experimental results indicate that the use of new parametric models allows the STC to improve the phase accuracy as well as the synthetic speech quality.
Text-independent speaker identification based on explicit exploitation of stochastic characteristics of test utterance
(2001-12-06)
Tsai, WH
;
Chang, WW
;
Huang, CS
;
電信工程研究所
;
Institute of Communications Engineering
The benefit of exploiting the stochastic characteristics of test utterance for speaker identification (speaker ID) via cross likelihood ratio and Bayesian information criterion is explored. Simulation results demonstrate the superiority of the proposed approaches over the conventional speaker ID based on maximum likelihood decision rule.
The automated cryptanalysis of DFT-based speech scramblers
(2000-12-01)
Chang, WW
;
Hsu, HI
;
電信工程研究所
;
Institute of Communications Engineering
An automated method For cryptanalysis of DFT-based analog speech scramblers is presented through statistical estimation treatments. In the proposed system, the cipher-text only attack is formulated as a combinatorial optimization problem leading to a search for the most likely key estimate. For greater efficiency, we also explore the benefits of genetic algorithm to develop an estimation method which takes into account the doubly stochastic characteristics of the underlying keyspace. Simulation results: indicate that the global explorative properties of genetic algorithms make them very effective at estimating the most likely permutation and by using this estimate significant amount of the intelligibility can be recovered from the ciphertext following the attack on DFT-based speech scramblers.