Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Bolbosh: Script-Aware Flow Matching for Kashmiri Text-to-Speech

Tajamul Ashraf, Burhaan Rasheed Zargar, Saeed Abdul Muizz, Ifrah Mushtaq, Nazima Mehdi, Iqra Altaf Gillani, Aadil Amin Kak, Janibul Bashir

TL;DR

Bolbosh is presented, a supervised cross-lingual adaptation strategy based on Optimal Transport Conditional Flow Matching (OT-CFM) within the Matcha-TTS framework that enables stable alignment under limited paired data.

Abstract

Kashmiri is spoken by around 7 million people but remains critically underserved in speech technology, despite its official status and rich linguistic heritage. The lack of robust Text-to-Speech (TTS) systems limits digital accessibility and inclusive human-computer interaction for native speakers. In this work, we present the first dedicated open-source neural TTS system designed for Kashmiri. We show that zero-shot multilingual baselines trained for Indic languages fail to produce intelligible speech, achieving a Mean Opinion Score (MOS) of only 1.86, largely due to inadequate modeling of Perso-Arabic diacritics and language-specific phonotactics. To address these limitations, we propose Bolbosh, a supervised cross-lingual adaptation strategy based on Optimal Transport Conditional Flow Matching (OT-CFM) within the Matcha-TTS framework. This enables stable alignment under limited paired data. We further introduce a three-stage acoustic enhancement pipeline consisting of dereverberation, silence trimming, and loudness normalization to unify heterogeneous speech sources and stabilize alignment learning. The model vocabulary is expanded to explicitly encode Kashmiri graphemes, preserving fine-grained vowel distinctions. Our system achieves a MOS of 3.63 and a Mel-Cepstral Distortion (MCD) of 3.73, substantially outperforming multilingual baselines and establishing a new benchmark for Kashmiri speech synthesis. Our results demonstrate that script-aware and supervised flow-based adaptation are critical for low-resource TTS in diacritic-sensitive languages. Code and data are available at: https://github.com/gaash-lab/Bolbosh.

Bolbosh: Script-Aware Flow Matching for Kashmiri Text-to-Speech

TL;DR

Bolbosh is presented, a supervised cross-lingual adaptation strategy based on Optimal Transport Conditional Flow Matching (OT-CFM) within the Matcha-TTS framework that enables stable alignment under limited paired data.

Abstract

Kashmiri is spoken by around 7 million people but remains critically underserved in speech technology, despite its official status and rich linguistic heritage. The lack of robust Text-to-Speech (TTS) systems limits digital accessibility and inclusive human-computer interaction for native speakers. In this work, we present the first dedicated open-source neural TTS system designed for Kashmiri. We show that zero-shot multilingual baselines trained for Indic languages fail to produce intelligible speech, achieving a Mean Opinion Score (MOS) of only 1.86, largely due to inadequate modeling of Perso-Arabic diacritics and language-specific phonotactics. To address these limitations, we propose Bolbosh, a supervised cross-lingual adaptation strategy based on Optimal Transport Conditional Flow Matching (OT-CFM) within the Matcha-TTS framework. This enables stable alignment under limited paired data. We further introduce a three-stage acoustic enhancement pipeline consisting of dereverberation, silence trimming, and loudness normalization to unify heterogeneous speech sources and stabilize alignment learning. The model vocabulary is expanded to explicitly encode Kashmiri graphemes, preserving fine-grained vowel distinctions. Our system achieves a MOS of 3.63 and a Mel-Cepstral Distortion (MCD) of 3.73, substantially outperforming multilingual baselines and establishing a new benchmark for Kashmiri speech synthesis. Our results demonstrate that script-aware and supervised flow-based adaptation are critical for low-resource TTS in diacritic-sensitive languages. Code and data are available at: https://github.com/gaash-lab/Bolbosh.
Paper Structure (20 sections, 1 equation, 1 figure, 4 tables)

This paper contains 20 sections, 1 equation, 1 figure, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Speech Technology for Indic Languages
  4. Flow-Matching Based TTS
  5. Kashmiri TTS Dataset
  6. Dataset
  7. IndicVoices-R Enhancement Pipeline
  8. Text Processing and Normalization
  9. Proposed Framework
  10. Model Architecture
  11. Training Objective
  12. Low-Resource Adaptation Strategy
  13. Results and Discussion
  14. Implementation Details
  15. Evaluation Metrics
  16. ...and 5 more sections

Figures (1)

  • Figure 1: Mel-spectrogram comparison of 2 Kashmiri utterances. The utterances were synthesized for the following text: Right (IPA): [me ru:d n1 pA:n2s tA:m jel1 lUkow t2th @nd2s p2kn1 bA:p@th s2kI pj2h træphIk A:m@s m2nz bA:kIjow khot1 brõh nj@:rn1 kh@:tr1 g@:j2l@:w j2th pj2h b1 p@kA:n o:sUs] Left (IPA): [me h1 khoSi: zI seh2t m2rk@z@s m2nz j1m h2pht1 kIs @:khr@s pj2h m2Sw@r1 kh@:tr1 wA:rj@h t2rUb1 kA:r A:kh@r d2stIjA:b]