Building Statistical Parametric Multi-speaker Synthesis for Bangladeshi Bangla

Gutkin, Alexander; Ha, Linne; Jansche, Martin; Kjartansson, Oddur; Pipatsrisawat, Knot; Sproat, Richard

doi:10.1016/j.procs.2016.04.049

Cited by 7 publications

(6 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The corpus has both male and female speakers and is quite mixed: Most datasets are single-speaker, while others, like Bangladeshi Bengali and Icelandic, have recordings from multiple speakers [9]. The recording conditions vary: some speakers were recorded in anechoic chambers, while others in a regular recording studio setup or on university campuses.…”

Section: Methodsmentioning

confidence: 99%

“…However, this approach may not apply for under-resourced languages, when no source corpora are available. In this scenario, crowd-sourcing the data from multiple speakers and building an average voice is possible [9]. For the majority of the world's languages, in the long tail of the distribution [10], even these approaches may not be feasible, due to the lack of sufficient audio data, linguistic resources, or adequate infrastructure [11].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Uniform Multilingual Multi-Speaker Acoustic Model for Statistical Parametric Speech Synthesis of Low-Resourced Languages

Gutkin

2017

Interspeech 2017

Self Cite

View full text Add to dashboard Cite

Acquiring data for text-to-speech (TTS) systems is expensive. This typically requires large amounts of training data, which is not available for low-resourced languages. Sometimes small amounts of data can be collected, while often no data may be available at all. This paper presents an acoustic modeling approach utilizing long short-term memory (LSTM) recurrent neural networks (RNN) aimed at partially addressing the language data scarcity problem. Unlike speaker-adaptation systems that aim to preserve speaker similarity across languages, the salient feature of the proposed approach is that, once constructed, the resulting system does not need retraining to cope with the previously unseen languages. This is due to language and speaker-agnostic model topology and universal linguistic feature set. Experiments on twelve languages show that the system is able to produce intelligible and sometimes natural output when a language is unseen. We also show that, when small amounts of training data are available, pooling the data sometimes improves the overall intelligibility and naturalness. Finally, we show that sometimes having a multilingual system with no prior exposure to the language is better than building single-speaker system from small amounts of data for that language. Index Terms: speech synthesis, low-resourced languages, long short-term memory, recurrent neural networks 2.1. Linguistic features Typically, the input to the acoustic model in statistical parametric speech synthesis consists of many diverse types of linguistic features. The features may include positional information (po

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Uniform Multilingual Multi-Speaker Acoustic Model for Statistical Parametric Speech Synthesis of Low-Resourced Languages

Gutkin

2017

Interspeech 2017

Self Cite

View full text Add to dashboard Cite

show abstract

“…There have been numerous attempts at solving this data bottleneck. Text-to-speech voices have been successfully trained on crowdsourced speech data [2], on filtered acoustic data for training speech recognition systems [3], or commercial audiobooks [4]. New voices have also been created using voice adaptation techniques [5] or voice morphing [6].…”

Section: Introductionmentioning

confidence: 99%

Experiments with Training Corpora for Statistical Text-to-speech Systems.

Podsiadło¹,

Ungureanu²

2018

Interspeech 2018

View full text Add to dashboard Cite

Common text-to-speech (TTS) systems rely on training data for modelling human speech. The quality of this data can range from professional voice actors recording hand-curated sentences in high-quality studio conditions, to found voice data representing arbitrary domains. For years, the unit selection technology dominant in the field required many hours of data that was expensive and time-consuming to collect. With the advancement of statistical methods of waveform generation, there have been experiments with more noisy and often much larger datasets, testing the inherent flexibility of such systems. In this paper we examine the relationship between training data and speech synthesis quality. We then hypothesise that statistical text-to-speech benefits from high acoustic quality corpora with high level of prosodic variation, but that beyond the first few hours of training data we do not observe quality gains. We then describe how we engineered a training dataset containing optimized distribution of features, and how these features were defined. Lastly, we present results from a series of evaluation tests. These confirm our hypothesis and show how a carefully engineered training corpus of a smaller size yields the same speech quality as much larger datasets, particularly for voices that use WaveNet.

show abstract

“…Recent work has been successful at developing text-to-speech (TTS) resources and systems for under-resourced languages by combining inexpensive methods such as crowd-sourcing and bootstrapping with powerful statistical acoustic modelling techniques [1]. The South African (SA) context with 11 official languages, most of them under-resourced but some closely related, provides an interesting scenario for further investigating rapid development of practical systems with similar efficient techniques.…”

Section: Introductionmentioning

confidence: 99%

Rapid Development of TTS Corpora for Four South African Languages

Niekerk¹,

Heerden²,

Davel³

et al. 2017

Interspeech 2017

Self Cite

View full text Add to dashboard Cite

This paper describes the development of text-to-speech corpora for four South African languages. The approach followed investigated the possibility of using low-cost methods including informal recording environments and untrained volunteer speakers. This objective and the additional future goal of expanding the corpus to increase coverage of South Africa's 11 official languages necessitated experimenting with multi-speaker and code-switched data. The process and relevant observations are detailed throughout. The latest version of the corpora are available for download under an open-source licence and will likely see further development and refinement in future.

show abstract

Building Statistical Parametric Multi-speaker Synthesis for Bangladeshi Bangla

Cited by 7 publications

References 22 publications

Uniform Multilingual Multi-Speaker Acoustic Model for Statistical Parametric Speech Synthesis of Low-Resourced Languages

Uniform Multilingual Multi-Speaker Acoustic Model for Statistical Parametric Speech Synthesis of Low-Resourced Languages

Experiments with Training Corpora for Statistical Text-to-speech Systems.

Rapid Development of TTS Corpora for Four South African Languages

Contact Info

Product

Resources

About