mlsanityData to 14 Oct 2026

ManchuTTS: Towards High-Quality Manchu Speech Synthesis via Flow Matching and Hierarchical Text Representation

arXiv 2512.22491 ↗First posted 27 Dec 2025cs.CL cs.AIcited by 0

On the map: Diffusion models → Audio, speech and music

Abstract (arXiv)

As an endangered language, Manchu presents unique challenges for speech synthesis, including severe data scarcity and strong phonological agglutination. This paper proposes ManchuTTS(Manchu Text to Speech), a novel approach tailored to Manchu's linguistic characteristics. To handle agglutination, this method designs a three-tier text representation (phoneme, syllable, prosodic) and a cross-modal hierarchical attention mechanism for multi-granular alignment. The synthesis model integrates deep convolutional networks with a flow-matching Transformer, enabling efficient, non-autoregressive generation. This method further introduce a hierarchical contrastive loss to guide structured acoustic-linguistic correspondence. To address low-resource constraints, This method construct the first Manchu TTS dataset and employ a data augmentation strategy. Experiments demonstrate that ManchuTTS attains a MOS of 4.52 using a 5.2-hour training subset derived from our full 6.24-hour annotated corpus, outperforming all baseline models by a notable margin. Ablations confirm hierarchical guidance improves agglutinative word pronunciation accuracy (AWPA) by 31% and prosodic naturalness by 27%.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Audio, speech and music, one of 14 directions in Diffusion models. Builds on 6 of its 13 references that are in the map. First posted on arXiv 27 Dec 2025.

Builds on

6 of 13 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025SlimSpeech: Lightweight and Efficient Text-to-Speech with Slim Rectified Flow10 Apr
  2. 2025Long-Term TalkingFace Generation via Motion-Prior Conditional Diffusion Model13 Feb
  3. 2024F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching9 Oct

Other references · 3

  1. 2020Glow-TTS: A Generative Flow for Text-to-Speech via Monotonic Alignment Searchcited by 273
  2. 2024Transcending Language Boundaries: Harnessing LLMs for Low-Resource Language Translationcited by 12
  3. 2024Wave-U-Mamba: An End-To-End Framework For High-Quality And Efficient Speech Super Resolutioncited by 9

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025DiaMoE-TTS: A Unified IPA-Based Dialect TTS Framework with Mixture-of-Experts and Parameter-Efficient Zero-Shot Adaptation
  2. 2026PolySpeech-100: A Large-Scale Benchmark for Speech Understanding Across 100+ Languages and Dialects
  3. 2025Developing multilingual speech synthesis system for Ojibwe, Mi'kmaq, and Maliseet
  4. 2026OmniVoice: Towards Omnilingual Zero-Shot Text-to-Speech with Diffusion Language Models
  5. 2023A Vector Quantized Approach for Text to Speech Synthesis on Real-World Spontaneous Speech
  6. 2026CuteTTS: Efficient and High-Quality Speech Synthesis via Autoregressive Modeling of Continuous Latents
  7. 2025MegaTTS 3: Sparse Alignment Enhanced Latent Diffusion Transformer for Zero-Shot Speech Synthesis
  8. 2026FlowEdit: Associative Memory for Lifelong Pronunciation Adaptation in Flow-Matching TTS