Towards automatic cross-lingual acoustic modelling applied to HMM-based speech synthesis for under-resourced languages

Abstract

Original scientific paperNowadays Human Computer Interaction (HCI) can also be achieved with voice user interfaces (VUIs). To enable devices to communicate with humans by speech in the user's own language, low-cost language portability is often discussed and analysed. One of the most time-consuming parts for the language-adaptation process of VUIcapable applications is the target-language speech-data acquisition. Such data is further used in the development of VUIs subsystems, especially of speech-recognition and speech-production systems. The tempting idea to bypass a long-term process of data acquisition is considering the design and development of an automatic algorithms, which can extract the similar target-language acoustic from different language speech databases. This paper focus on the cross-lingual phoneme mapping between an under-resourced and a well-resourced language. It proposes a novel automatic phoneme-mapping technique that is adopted from the speaker-verification field. Such a phoneme mapping is further used in the development of the HMM-based speech-synthesis system for the under-resourced language. The synthesised utterances are evaluated with a subjective evaluation and compared by the expert knowledge cross-language method against to the baseline speech synthesis based just from the under-resourced data. The results reveals, that combining data from well-resourced and under-resourced language with the use of the proposed phoneme-mapping technique, can improve the quality of under-resourced language speech synthesis.Key words: Voice user interfaces, Human language technologies, HMM-based speech synthesis, Cross-language synthesis, Under-resourced languages, UBM-MAP-GMM phoneme mapping Primjena automatskog medujezičnog akustičnog modeliranja na HMM sintezu govora za oskudne jezične baze. U današnje vrijeme interakcijačovjeka i računala (HCI) može se ostvariti i putem govornih sučelja (VUIs). Da bi se omogućila komunikacija uredaja i korisnika putem govora na vlastitom korisnikovom jeziku, cesto se raspravlja i analizira o jeftinom rješenju prijevoda govora na različite jezike. Jedan od vremenski najzahtjevnijih dijelova procesa prilagodbe jezika za aplikacije koje podržavaju VUI je prikupljanje govornih podataka za ciljani jezik. Ovakvi podaci dalje se koriste za razvoj VUI podsustava, posebice za prepoznavanje i produkciju govora. Primamljiva ideja za izbjegavanje dugotrajnog postupka prikupljanja podataka jeste razmatranje sinteze i razvoja automatskih algoritama koji su sposobni izvesti slična akustična svojstva za ciljani jezik iz postojećih baza različitih jezika. Ovaj rad fokusiran je na povezivanje medujezičnih fonema izmedu oskudnih i bogatih jezičnih baza. Predložena je nova tehnika automatskog povezivanja fonema, usvojena i prilagodena iz područja govorne autentikacije. Ovakvo povezivanje fonema kasnije se koristi za razvoj sustava za sintezu govora zasnovanom na HMM-u za manje poznate jezike. Načinjene govorne izjave ocijenjene su subjektivnim pristupom kroz usporedbu medujezičnih m...

A Systematic Review and Analysis of Multilingual Data Strategies in Text-to-Speech for Low-Resource Languages

“…These resulting values (n = 880) were used for analysis. [6], [7], [8], [9], [10], [11], [12] Hidden Markov Model synthesis (HMM) 7 [12], [13], [14], [15], [16], [17], [18] Neural network (non-S2S) synthesis (DNN) 9 [19], [20], [21], [22], [23], [24], [25], [26], [27] Sequence-to-sequence synthesis (S2S)…”

“…Intelligibility [23] WER (Word Error Rate) Intelligibility [20] MOS (Mean Opinion Score) Naturalness/ Quality [6], [8], [9], [14], [15], [16], [17], [18], [21], [23] A/B Preference (preference rate b/w test & control) Quality [5], [10], [11], [12], [13]…”

“…Since all studies (except for [19] and [26]) mentioned this only in either time duration or number of utterances, an average estimation was needed to convert it to a common measurement. Following the descriptions provided in [10], [11], [12], [16], [17], [20], [21], [22], [23], [24], and [27], we obtained an average utterance length of the speech data sets used in these studies: 6.1 seconds. This was then used to convert all the training data quantities to the corresponding estimated number of utterances.…”