mlsanityData to 19 Oct 2026

Vocoders and speech synthesis

1,631 papers2 this weekMedium confidence in this grouping: mostly one topic, with blurry edgesMedium confidence in this grouping: mostly one topic, with blurry edges · About

Keywords: speech, tts, synthesis, text-to-speech, audio

Key facts. Vocoders and speech synthesis: 1,631 papers on arXiv since 2021 (data through 19 Oct 2026). 486 of its 1,631 mapped papers (30%) cite “Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers” (2023, arXiv 2301.02111). 28 new papers 2 Oct–19 Oct.

What’s new

2 Oct – 19 Oct 2026

28 new papers, 2 Oct – 19 Oct; 2 this week.

  1. This week · 2
  2. Steering intonational tunes in text-to-speech synthesis also fits: Audio language models and spoken dialogue
  3. SEAM: Spectral Embedding Alignment and Merging for Low-Resource TTS
  4. Earlier · 26
  5. MixiSpeech: Progressive State Distillation for One-Step Vietnamese TTS
  6. ShlokaDhara: A Multi-Speaker Sanskrit Speech Corpus for Learning Meter-Aware Prosody in Text-to-Recitation
  7. Beyond the Benchmark: Conditioning, Domains and Failure Modes of Qwen3-TTS
  8. CoSTA: Compressed Speech Token Adaptation for Efficient Spoken Dialogue Models also fits: Speech translation
  9. Probing Neural Audio Codecs with Speaker Traits and Speech Characteristics
  10. Revise-TTS: A Non-Masking Refinement Framework for Masked Diffusion Models in Text-to-Speech
  11. PACT: A Pragmatic Prompting Framework for Expressive and Voice-Preserving Speech-to-Speech Translation also fits: Speech translation
  12. BridgeCodec: Mamba Enhanced Neural Audio Codec with Schr"odinger Bridge at Low Bitrate also fits: Speech enhancement and noise suppression
All 28
  1. Training a Compact, Strong, and Realistic Text-to-Speech Model
  2. VoTSpeech: Decoupling Voice Design from Speech Generation in Instruction-Following TTS
  3. AudioTETRIS: Token Elimination via Transitional Importance Signals for Efficient Audio LLMs also fits: Audio language models and spoken dialogue
  4. Beyond Speech Captions: Speech-Rewarded Style Planning for Conversational Text-to-Speech also fits: Audio language models and spoken dialogue
  5. Edit Who Speaks, Control How They Speak: Global Timbre Editing and Local Instruction Control for TTS
  6. Local Prototype Reconstruction for Text-Compatible Speech-to-LLM Bridge Pretraining also fits: Speech translation
  7. Training-Free Instruction TTS Gender Bias Calibration Using Model-Adaptive Steering
  8. Beyond Token Revision: Investigating Mask-and-Replace Diffusion for Zero-Shot Text-to-Speech
  9. Hearing's Blind Spots: Toward Perceptual Compute Allocation in Neural Text-to-Speech also fits: Speech enhancement and noise suppression
  10. Loud and Clear: Dynamic Activation Steering for Improving Speech Intelligibility in Noisy Environments also fits: Audio language models and spoken dialogue
  11. Pronunciation-Oriented Reinforcement Learning for Japanese Text-to-Speech with Kana-Domain ASR Rewards also fits: Multilingual and low-resource ASR
  12. Region-Aware Masking for Accent-Robust Cross-Lingual Text-to-Speech also fits: Speech translation
  13. Neural Representations, Natural Connections: What Transfers From Human Speech Foundation Models to Animal Vocalizations? also fits: Audio language models and spoken dialogue
  14. Paradee: Distilling Kokoro-82M into an 8M-Parameter Single-Voice Text-to-Speech Model also fits: Audio language models and spoken dialogue
  15. A Comprehensive Objective Evaluation of Modern Text-to-Speech for Turkish Using Speech Quality Assessment Models also fits: Multilingual and low-resource ASR
  16. GS-Codec: A Gaussian-Splatting Bottleneck for Neural Audio Coding
  17. DriftTTS: Few-Step Text-to-Speech Without Distillation via Distribution-Matching Drift
  18. Refinement Buys Intelligibility, Search Buys Identity: What Test-Time Compute Buys in Masked-Diffusion TTS also fits: Audio language models and spoken dialogue

Foundations

Most cited by papers here

  1. Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers2023cited by 486 here
  2. SoundStream: An End-to-End Neural Audio Codec2021cited by 358 here
  3. Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech2021cited by 344 here
  4. F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching2024cited by 242 here
  5. Voicebox: Text-Guided Multilingual Universal Speech Generation at Scale2023cited by 231 here

Lineage

The papers this direction builds on, oldest first.

Each paper on the line cites the one before.

  1. 2021Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speechcited by 344 here
  2. 2021A Survey on Neural Speech Synthesiscited by 124 here
  3. 2021YourTTS: Towards Zero-Shot Multi-Speaker TTS and Zero-Shot Voice Conversion for everyonecited by 193 here
  4. 2023Neural Codec Language Models are Zero-Shot Text to Speech Synthesizerscited by 486 here
  5. 2023Speak, Read and Prompt: High-Fidelity Text-to-Speech with Minimal Supervisioncited by 157 here
  6. 2023NaturalSpeech 2: Latent Diffusion Models are Natural and Zero-Shot Speech and Singing Synthesizerscited by 163 here
  7. 2023Voicebox: Text-Guided Multilingual Universal Speech Generation at Scalecited by 231 here
  8. 2024NaturalSpeech 3: Zero-Shot Speech Synthesis with Factorized Codec and Diffusion Modelscited by 185 here
  9. 2024VALL-E 2: Neural Codec Language Models are Human Parity Zero-Shot Text to Speech Synthesizerscited by 106 here
  10. 2024E2 TTS: Embarrassingly Easy Fully Non-Autoregressive Zero-Shot TTScited by 137 here
  11. 2024MaskGCT: Zero-Shot Text-to-Speech with Masked Generative Codec Transformercited by 137 here
  12. 2024F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matchingcited by 242 here
  13. 2024CosyVoice 2: Scalable Streaming Speech Synthesis with Large Language Modelscited by 164 here
11 more it builds on, not in this chain
  1. 2021Generative Spoken Language Modeling from Raw Audiocited by 97 here
  2. 2021Grad-TTS: A Diffusion Probabilistic Model for Text-to-Speechcited by 166 here
  3. 2021SoundStream: An End-to-End Neural Audio Codeccited by 358 here
  4. 2021WavLM: Large-Scale Self-Supervised Pre-Training for Full Stack Speech Processingcited by 347 here
  5. 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 460 here
  6. 2023Speak Foreign Languages with Your Own Voice: Cross-Lingual Neural Codec Language Modelingcited by 109 here
  7. 2023Matcha-TTS: A fast TTS architecture with conditional flow matchingcited by 100 here
  8. 2024Seed-TTS: A Family of High-Quality Versatile Speech Generation Modelscited by 208 here
  9. 2024CosyVoice: A Scalable Multilingual Zero-shot Text-to-speech Synthesizer based on Supervised Semantic Tokenscited by 209 here
  10. 2024WavTokenizer: an Efficient Acoustic Discrete Codec Tokenizer for Audio Language Modelingcited by 98 here
  11. 2024Moshi: a speech-text foundation model for real-time dialoguecited by 153 here

The 24 most cited of 2,021.

Browse

30 of 1,631, ranked 6–35

  1. 6CosyVoice: A Scalable Multilingual Zero-shot Text-to-speech Synthesizer based on Supervised Semantic Tokens2024209
  2. 7Seed-TTS: A Family of High-Quality Versatile Speech Generation Models2024208
  3. 8YourTTS: Towards Zero-Shot Multi-Speaker TTS and Zero-Shot Voice Conversion for everyone2021193
  4. 9NaturalSpeech 3: Zero-Shot Speech Synthesis with Factorized Codec and Diffusion Models2024185
  5. 10Grad-TTS: A Diffusion Probabilistic Model for Text-to-Speech2021166
  6. 11CosyVoice 2: Scalable Streaming Speech Synthesis with Large Language Models2024164
  7. 12NaturalSpeech 2: Latent Diffusion Models are Natural and Zero-Shot Speech and Singing Synthesizers2023163
  8. 13Speak, Read and Prompt: High-Fidelity Text-to-Speech with Minimal Supervision2023157
  9. 14E2 TTS: Embarrassingly Easy Fully Non-Autoregressive Zero-Shot TTS2024137
  10. 15MaskGCT: Zero-Shot Text-to-Speech with Masked Generative Codec Transformer2024137
All 30
  1. 16A Survey on Neural Speech Synthesis2021124
  2. 17Speak Foreign Languages with Your Own Voice: Cross-Lingual Neural Codec Language Modeling2023109
  3. 18VALL-E 2: Neural Codec Language Models are Human Parity Zero-Shot Text to Speech Synthesizers2024106
  4. 19Matcha-TTS: A fast TTS architecture with conditional flow matching2023100
  5. 20WavTokenizer: an Efficient Acoustic Discrete Codec Tokenizer for Audio Language Modeling202498
  6. 21Generative Spoken Language Modeling from Raw Audio202197
  7. 22XTTS: a Massively Multilingual Zero-Shot Text-to-Speech Model202496
  8. 23CosyVoice 3: Towards In-the-wild Speech Generation via Scaling-up and Post-training202594
  9. 24Spark-TTS: An Efficient LLM-Based Text-to-Speech Model with Single-Stream Decoupled Speech Tokens202591
  10. 25NaturalSpeech: End-to-End Text to Speech Synthesis with Human-Level Quality202289
  11. 26VoiceCraft: Zero-Shot Speech Editing and Text-to-Speech in the Wild202488
  12. 27HiFi-Codec: Group-residual Vector quantization for High Fidelity Audio Codec202388
  13. 28StyleTTS 2: Towards Human-Level Text-to-Speech through Style Diffusion and Adversarial Training with Large Speech Language Models202383
  14. 29SpeechTokenizer: Unified Speech Tokenizer for Speech Large Language Models202383
  15. 30AdaSpeech: Adaptive Text to Speech for Custom Voice202180
  16. 31Meta-StyleSpeech : Multi-Speaker Adaptive Text-to-Speech Generation202176
  17. 32Diff-TTS: A Denoising Diffusion Model for Text-to-Speech202172
  18. 33FireRedTTS: A Foundation Text-To-Speech Framework for Industry-Level Generative Speech Applications202471
  19. 34Llasa: Scaling Train-Time and Inference-Time Compute for Llama-based Speech Synthesis202570
  20. 35LibriTTS-R: A Restored Multi-Speaker Text-to-Speech Corpus202370