<i>Speech Processing and Synthesis Toolboxes</i>

Childers, D. G.; Díaz, José

doi:10.1121/1.1318896

Cited by 106 publications

(99 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The LF parameters used for generating the glottal source are specified in per cent of the duration of the glottal cycles as listed in Table IV. In addition, the LF parameters obey a normal distribution with standard deviation of 2 % around these nominal values, varying from glottal cycle to glottal cycle, as described in [30]. The results for each experiment and each test configuration were averaged from 100 glottal cycles.…”

Section: A Methodology Validation Using Synthetic Speechmentioning

confidence: 99%

“…(4) ensures that the glottal flow waveform returns to zero after each glottal cycle and is typically enforced by iteratively optimizing the damping parameter α of the exponential segment in Eq. (4) [30].…”

Section: B the Liljencrants-fant Voice Source Modelmentioning

confidence: 99%

“…The boundary constraints for the parameters are listed in Table I and Table II. The values for the formant frequencies were derived from [30] and the values for the formant bandwidths were taken from [39].…”

Section: B Pitch-synchronous Optimizationmentioning

confidence: 99%

See 2 more Smart Citations

Joint Source-Filter Optimization for Accurate Vocal Tract Estimation Using Differential Evolution

Schleusing

Kinnunen

Story

et al. 2013

IEEE Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

Abstract-In this work, we present a joint source-filter optimization approach for separating voiced speech into vocal tract (VT) and voice source components. The presented method is pitch-synchronous and thereby exhibits a high robustness against vocal jitter, shimmer and other glottal variations while covering various voice qualities. The voice source is modeled using the Liljencrants-Fant (LF) model, which is integrated into a time-varying auto-regressive speech production model with exogenous input (ARX). The non-convex optimization problem of finding the optimal model parameters is addressed by a heuristic, evolutionary optimization method called differential evolution. The optimization method is first validated in a series of experiments with synthetic speech. Estimated glottal source and VT parameters are the criteria used for comparison with the iterative adaptive inverse filter (IAIF) method and the linear prediction (LP) method under varying conditions such as jitter, fundamental frequency (f0) as well as environmental and glottal noise. The results show that the proposed method largely reduces the bias and standard deviation of estimated VT coefficients and glottal source parameters. Furthermore, the performance of the source-filter separation is evaluated in experiments using speech generated with a physical model of speech production. The proposed method reliably estimates glottal flow waveforms and lower formant frequencies. Results obtained for higher formant frequencies indicate that research on more accurate voice source models and their interaction with the VT is necessary to improve the source-filter separation. The proposed optimization approach promises to be a useful tool for future research addressing this topic.

show abstract

Section: A Methodology Validation Using Synthetic Speechmentioning

confidence: 99%

Section: B the Liljencrants-fant Voice Source Modelmentioning

confidence: 99%

See 1 more Smart Citation

Joint Source-Filter Optimization for Accurate Vocal Tract Estimation Using Differential Evolution

Schleusing

Kinnunen

Story

et al. 2013

IEEE Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

show abstract

“…Figure 1 shows the main components of a typical conversational system (Zue and Glass 2000). These systems can be speaker dependant as in the case of dictation products or speaker independent in the case customer service applications with large number of callers (Childers 2000). The move to provide Speech services via the Web is being facilitated by VXML (Voice eXtensible Mark-up Language), a standard of the World Wide Web Consortium (W3C 2005), which allows conventional telephony interfaces such as IVR to evolve using voice gateways linked to a Web server.…”

Section: Speech and Language Processingmentioning

confidence: 99%

The Diffusion of WOZ: Expanding the Topology of IS Innovations

Costello

Donnellan

2007

Journal of Information Technology

View full text Add to dashboard Cite

The growth and diffusion of self-service technology (SST) over the last decade has resulted in an increasing number of business and government transactions being completed without human assistance. One innovation in this area, the speech-enabled business system, is characterised by complex implementations that bring together language processing technologies, applications development and end-user psychology. A resulting secondary innovation, the Wizard-of-Oz (WOZ) experiment is a valuable technique for simulating and building human-machine prototypes to ensure successful deployment of the completed service. The objective of the paper is to examine these innovations in relation to the changing business landscape; the technology and innovation literature, and the population of likely adopters. The review is carried out by placing the author's former experience as a practitioner within current theoretical research frameworks. The result is a number of suggestions relating to both IT technology research and IT innovation research. Firstly, it proposes the simplification and diffusion of the WOZ methodology to support the growth in demand for automated e-business transactions that is mindful of human and ethical challenges. Secondly, the paper argues that because self-service technology and business extends the traditional boundaries of the customer service function, it now needs to be incorporated into Swanson's tri-core innovation typology. The paper concludes by presenting the suggested reorientation of IS research that incorporates an outward facing perceptive as a conceptual model.

show abstract

“…A general score function is defined in equation (7) based on the method described in [13]. In other words the score function evaluates values of the vector x at indexes defined by the vector idx against a lower threshold c 1 and an upper threshold c 2 , returning scores in the range 0 to 1 at each index.…”

Section: C) Candidate Boundary Scoringmentioning

confidence: 99%

Automatic blind syllable segmentation for continuous speech

Villing

Timoney²,

Ward³

2004

Irish Signals and Systems Conference 2004

View full text Add to dashboard Cite

--In this paper a simple practical method for blind segmentation of continuous speech into its constituent syllables is presented. This technique which uses amplitude onset velocity and coarse spectral makeup to identify syllable boundaries is tested on a corpus of continuous speech and compared with an established segmentation algorithm. The results show substantial performance benefit using the proposed algorithm.

show abstract

Speech Processing and Synthesis Toolboxes

Cited by 106 publications

References 0 publications

Joint Source-Filter Optimization for Accurate Vocal Tract Estimation Using Differential Evolution

Joint Source-Filter Optimization for Accurate Vocal Tract Estimation Using Differential Evolution

The Diffusion of WOZ: Expanding the Topology of IS Innovations

Automatic blind syllable segmentation for continuous speech

Contact Info

Product

Resources

About