Vocaine the vocoder and applications in speech synthesis

Agiomyrgiannakis, Yannis

doi:10.1109/icassp.2015.7178768

Cited by 59 publications

(49 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ODFWW algorithm is evaluated in the context of a high-quality Vocaine+LSTM-based TTS (VL-TTS) synthesizer [2]. Spectral and log(f0) modifications are made during runtime in the vocoder.…”

Section: Resultsmentioning

confidence: 99%

“…The overall increase in complexity in the vocoder is minimized to a modest +8% in a benchmark Android/ARM device by computing the spectral warping and weighting and the power normalization together with the spectral sampling because the two processes share a lot of computations. Following [2], the speech parameterization consists of 40-dimensional MCEP (including power), log f0 and a 7-dimensional band aperiodicity. The target conversion data are derived from the CSTR VCTK corpus [32].…”

Section: Resultsmentioning

confidence: 99%

“…The frequency warping and weighting transform is used on the vocoder side of a statistical parametric TTS synthesizer [2] during spectral sampling according to equation (5). In addition, we apply some power normalization to preserve the power of the original frame by multiplying each sampled harmonic amplitudeT (w(ω h )) by a factor…”

Section: Run-time Voice Morphing For Vocaine-based Ttsmentioning

confidence: 99%

“…Recent developments in TTS allow us to significantly improve the quality of the synthesized speech, namely by using an LSTM to predict the acoustic parameters from the linguistic ones [10] and Vocaine as the vocoder [2]. Neural-Network-based speech synthesis [11,12] can also benefit from adaptation [13,14].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Voice Morphing that improves TTS quality using an optimal dynamic frequency warping-and-weighting transform

Agiomyrgiannakis

Roupakia

2016

2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

Dynamic Frequency Warping (DFW) is widely used to align spectra of different speakers. It has long been argued that frequency warping captures inter-speaker differences but DFW practice always involves a tricky preprocessing part to remove spectral tilt. The DFW residual is successfully used in Voice Morphing to improve the quality and the similarity of synthesized speech but the estimation of the DFW residual remains largely heuristic and sub-optimal. This paper presents a dynamic programming algorithm that simultaneously estimates the Optimal Frequency Warping and Weighting transform (ODFWW) and therefore needs no preprocessing step and fine-tuning while source/target-speaker data are matched using the Matching-Minimization algorithm [1]. The transform is used to morph the output of a state-of-the-art Vocaine-based [2] TTS synthesizer in order to generate different voices in runtime with only +8% computational overhead. Some morphed TTS voices exhibit significantly higher quality than the original one as morphing seems to "correct" the voice characteristics of the TTS voice.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Run-time Voice Morphing For Vocaine-based Ttsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Voice Morphing that improves TTS quality using an optimal dynamic frequency warping-and-weighting transform

Agiomyrgiannakis

Roupakia

2016

2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Spectral enhancement based on postfiltering in the cepstral domain [39] was applied to improve the naturalness of the synthesized speech. From the acoustic features, speech waveforms were synthesized using the Vocaine vocoder [40].…”

Section: Experimental Conditionsmentioning

confidence: 99%

Unidirectional long short-term memory recurrent neural network with recurrent output layer for low-latency speech synthesis

Zen

Sak

2015

2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

247

197

View full text Add to dashboard Cite

Long short-term memory recurrent neural networks (LSTM-RNNs) have been applied to various speech applications including acoustic modeling for statistical parametric speech synthesis. One of the concerns for applying them to text-to-speech applications is its effect on latency. To address this concern, this paper proposes a low-latency, streaming speech synthesis architecture using unidirectional LSTMRNNs with a recurrent output layer. The use of unidirectional RNN architecture allows frame-synchronous streaming inference of output acoustic features given input linguistic features. The recurrent output layer further encourages smooth transition between acoustic features at consecutive frames. Experimental results in subjective listening tests show that the proposed architecture can synthesize natural sounding speech without requiring utterance-level batch processing.

show abstract

Introduction: Phase Processing, History

Mowlaee

2016

Single Channel Phase‐Aware Signal Processing in Speech Communication: Theory and Practice

View full text Add to dashboard Cite

Vocaine the vocoder and applications in speech synthesis

Cited by 59 publications

References 26 publications

Voice Morphing that improves TTS quality using an optimal dynamic frequency warping-and-weighting transform

Voice Morphing that improves TTS quality using an optimal dynamic frequency warping-and-weighting transform

Unidirectional long short-term memory recurrent neural network with recurrent output layer for low-latency speech synthesis

Introduction: Phase Processing, History

Contact Info

Product

Resources

About