mlsanityData to 14 Oct 2026

Balalaika: Data-Centric, Prosody-Aware Annotation Pipeline for Russian Speech

arXiv 2507.13563 ↗First posted 17 Jul 2025cs.CL cs.SD eess.AScited by 0

Not on the map. Why?

Abstract (arXiv)

We introduce Balalaika, an open-source, data-centric pipeline for processing audio and producing prosody-aware annotations. It combines semantic VAD for context-preserving segmentation, multi-ASR ensembling with ROVER consensus decoding, while retaining optional word-level timestamps, followed by automatic quality and speaker-purity filtering. The text is further enriched with punctuation restoration, lexical stress and "textipae/textipaHe" normalization, and IPA phonemes. Using Balalaika, we build a 5.1k-hour multi-source Russian corpus with rich annotations, and show consistent gains under equalized training budgets for both speech denoising and TTS; ablations confirm complementary benefits of stress and punctuation and improved synthesis with stricter MOS filtering. The datasets are publicly available at hrefhttps://huggingface.co/collections/lab260/balalaika-datasetunderlineHuggingFace

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 32 of its 70 references that are in the map. First posted on arXiv 17 Jul 2025.

Builds on

32 of 70 references.

References · 32

  1. 2019Common Voice: A Massively-Multilingual Speech Corpuscited by 1,384
  2. 2019LibriTTS: A Corpus Derived from LibriSpeech for Text-to-Speechcited by 649
  3. 2021XLS-R: Self-supervised Cross-lingual Speech Representation Learning at Scalecited by 601
  4. 2021Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speechcited by 561
  5. 2015MUSAN: A Music, Speech, and Noise Corpuscited by 425
  6. 2014Deep Speech: Scaling up end-to-end speech recognitioncited by 354
  7. 2021NISQA: A Deep CNN-Self-Attention Model for Multidimensional Speech Quality Prediction with Crowdsourced Datasetscited by 192
  8. 2023Powerset multi-class cross entropy loss for neural speaker diarizationcited by 111
  9. 2023LibriTTS-R: A Restored Multi-Speaker Text-to-Speech Corpuscited by 101
  10. 2024The T05 System for The VoiceMOS Challenge 2024: Transfer Learning from Deep Image Classifier to Naturalness MOS Prediction of High-Quality Synthetic Speechcited by 83
  11. 2020ViSQOL v3: An Open Source Production Ready Objective Speech and Audio Metriccited by 79
  12. 2024A Survey on Speech Large Language Models for Understandingcited by 71
20 more
  1. 2024An Investigation of Incorporating Mamba for Speech Enhancementcited by 62
  2. 2023MP-SENet: A Speech Enhancement Model with Parallel Denoising of Magnitude and Phase Spectracited by 52
  3. 2024WenetSpeech4TTS: A 12,800-hour Mandarin TTS Corpus for Large Speech Generation Model Benchmarkcited by 40
  4. 2023MossFormer2: Combining Transformer and RNN-Free Recurrent Network for Enhanced Time-Domain Monaural Speech Separationcited by 39
  5. 2021Deep Learning Based Assessment of Synthetic Speech Naturalnesscited by 35
  6. 2020Transformer based Grapheme-to-Phoneme Conversioncited by 32
  7. 2023DeepFilterNet: Perceptually Motivated Real-Time Speech Enhancementcited by 25
  8. 2023VITS2: Improving Quality and Efficiency of Single-Stage Text-to-Speech with Adversarial Learning and Architecture Designcited by 25
  9. 2025Emilia: A Large-Scale, Extensive, Multilingual, and Diverse Dataset for Speech Generationcited by 20
  10. 2025Data Quality Issues in Multilingual Speech Datasets: The Need for Sociolinguistic Awareness and Proactive Language Planningcited by 11
  11. 2024TouchTTS: An Embarrassingly Simple TTS Framework that Everyone Can Touchcited by 10
  12. 2021Golos: Russian Dataset for Speech Researchcited by 9
  13. 2024Large Language User Interfaces: Voice Interactive User Interfaces powered by LLMscited by 4
  14. 2019RUSLAN: Russian Spoken Language Corpus for Speech Synthesiscited by 3
  15. 2023Prosody Analysis of Audiobookscited by 3
  16. 2025Breaking the Transcription Bottleneck: Fine-tuning ASR Models for Extremely Low-Resource Fieldwork Languagescited by 3
  17. 2025VoiceRestore: Flow-Matching Transformers for Speech Recording Quality Restorationcited by 2
  18. 2025TTSOps: A Closed-Loop Corpus Optimization Framework for Training Multi-Speaker TTS Models from Dark Datacited by 2
  19. 2024Enhancing Crowdsourced Audio for Text-to-Speech Modelscited by 1
  20. 2025Universal-2-TF: Robust All-Neural Text Formatting for ASRcited by 1

Built on by

Not cited yet.