Papers for

multilingual tts system builders

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

Plan-to-speech failures linked to persona sentiment conflicts studied

When the Persona Contradicts the Line: Diagnosing Plan-to-Synthesis Failures in Chain-of-Thought TTS

Abstract: Chain-of-thought TTS systems rely on language-model planners to translate persona and context into vocal delivery, but it is unclear where this intent is lost when the requested affect conflicts with lexical sentiment. We introduce a bilingual Japanese--English persona--line conflict benchmark in which each item pairs one line with congruent and conflicting personas and pre-declares the expected acoustic direction. We probe this direction in the planner-generated delivery phrase, the planner-written emotion vector, the QwenEmotion-generated vector, and the rendered waveform. Planner outputs are usually direction-correct, while downstream agreement can be substantially lower. Bypassing QwenEmotion provides only modest gains over the full planner-controlled Base. Supervising QwenEmotion raises the direction rate to 75% in Japanese and 84% in English. A learned text-to-latent interface reaches the highest English rate of 89% and the highest Japanese arousal rate of 88%. These results show that learned repair of the plan-to-synthesis interface can recover counter-textual delivery, although its effect differs across languages and acoustic axes.

Mon 26 OctComputation and LanguageSound
The gist
Sometimes, text-to-speech systems struggle to express the right emotions when a character's personality conflicts with the sentiment of the words they say. The authors created a test in both Japanese and English to study these conflicts and where the system fails. They found that planners usually set the right emotional direction, but the final speech often doesn't match perfectly. By improving the way the system connects planning to speech generation, they increased the accuracy of emotional delivery. Their results vary based on language and different speech features.
Open → 2610.35008v1

BanglaTTS gets improved quality and long-form speech dataset

BanglaKontho: Closing the Long-Form Gap in Bangla Text-to-Speech

Abstract: Bangla, the seventh most spoken language in the world, remains under-resourced for neural text-to-speech. Public Bangla speech corpora are dominated by short read-prompt utterances collected for speech recognition, leaving long-form prosody and consistent single-speaker narration uncovered. We present BanglaKontho, a single-speaker Bangla TTS corpus of 20 hours derived from professional audiobook recordings: 7,050 segmented utterances with verified transcripts at 24 kHz. We also release a reusable Bangla text normalizer covering Bangladeshi-style digit grouping, currency and date expressions, Danda punctuation and Unicode normalization, together with the full preprocessing pipeline. An MB-iSTFT-VITS baseline trained from scratch reaches 9.5% WER and 4.46 naturalness MOS, against 16.0% and 3.16 for the same architecture retrained on the 12-hour IndicTTS-Bn corpus. The corpus is released openly under CC BY-NC 4.0.

Thu 24 SeptComputation and Language
The gist
Bangla is a widely spoken language but lacks good tools to turn written text into natural speech, especially for long passages. The authors provide a new collection of 20 hours of professionally recorded Bangla audiobook speech as a dataset for building such systems. They also offer software to clean and prepare Bangla text for speech synthesis. Their baseline system trained on this new data speaks more naturally and accurately compared to previous smaller datasets.
Open → 2609.29146v1