WavTTS is the first raw-waveform diffusion TTS model using DiT flow matching and multi-scale mel supervision that approaches SOTA latent zero-shot performance while beating prior end-to-end models.
hub Mixed citations
Qwen3-TTS Technical Report
Mixed citation behavior. Most common role is background (50%).
45
Pith papers citing it
Background
50% of classified citations
abstract
In this report, we present the Qwen3-TTS series, a family of advanced multilingual, controllable, robust, and streaming text-to-speech models. Qwen3-TTS supports state-of-the-art 3-second voice cloning and description-based control, allowing both the creation of entirely novel voices and fine-grained manipulation over the output speech. Trained on over 5 million hours of speech data spanning 10 languages, Qwen3-TTS adopts a dual-track LM architecture for real-time synthesis, coupled with two speech tokenizers: 1) Qwen-TTS-Tokenizer-25Hz is a single-codebook codec emphasizing semantic content, which offers seamlessly integration with Qwen-Audio and enables streaming waveform reconstruction via a block-wise DiT. 2) Qwen-TTS-Tokenizer-12Hz achieves extreme bitrate reduction and ultra-low-latency streaming, enabling immediate first-packet emission ($97\,\mathrm{ms}$) through its 12.5 Hz, 16-layer multi-codebook design and a lightweight causal ConvNet. Extensive experiments indicate state-of-the-art performance across diverse objective and subjective benchmark (e.g., TTS multilingual test set, InstructTTSEval, and our long speech test set). To facilitate community research and development, we release both tokenizers and models under the Apache 2.0 license.
hub tools
citation-role summary
background 3
baseline 2
method 1
citation-polarity summary
years
2026 45verdicts
UNVERDICTED 45representative citing papers
FlexiSLM is the first spoken language model supporting dynamic and controllable frame rates on speech input and output, outperforming fixed-rate 7B models at high quality and enabling faster inference at lower rates like 6.25 Hz.
JAVEdit-100k is the first large-scale dataset for instruction-guided joint audio-visual video editing, accompanied by JAVEditBench and the JAVEdit model that outperforms baselines on five of six metrics.
PlanAudio introduces a unified autoregressive LLM framework with semantic latent chain-of-thought for generating composite speech and sound audio from free-form text, plus a new benchmark.
GibbsTTS combines a training-free kinetic-optimal scheduler with finite-step moment correction in MI-DFM to deliver top naturalness and strong speaker similarity in zero-shot TTS.
VITA-QinYu is the first expressive end-to-end spoken language model supporting role-playing and singing alongside conversation, trained on 15.8K hours of data and outperforming prior models on expressiveness and conversational benchmarks.
Talker-T2AV achieves better lip-sync accuracy, video quality, and audio quality than dual-branch baselines by separating high-level shared autoregressive modeling from modality-specific low-level diffusion refinement in a joint audio-video generation framework.
MINT-Bench is a new benchmark using hierarchical taxonomy, multi-stage data pipeline, and hybrid evaluation to assess instruction-following TTS systems, revealing major gaps in compositional and paralinguistic controls.
NVBench provides a standardized bilingual benchmark and evaluation protocol for assessing non-verbal vocalization generation, placement, and salience in text-to-speech systems.
HumDial-EIBench is a new benchmark using real human dialogues to evaluate audio language models on emotional intelligence tasks including multi-turn tracking, causal reasoning, empathy generation, and acoustic-semantic conflict resolution.
CAST benchmark shows language models infer correct word stress from discourse context but TTS systems frequently fail to produce it in speech.
CapTalk unifies single-utterance and dialogue voice design via utterance- and speaker-level captions plus a hierarchical variational module for stable timbre with adaptive expression.
UniSAE unifies speaker, emotion, and multi-granularity content editing in speech via a new discrete phonetic posteriorgram representation and diffusion-based rendering.
HPRO uses a differentiable HD-Emo codec to extract separate content and style tokens and progressively aligns frame-, word-, and sentence-level rewards to improve emotional expressiveness in TTS while preserving intelligibility.
Controllable neural TTS reveals that loudness drives human sarcasm perception while a model prioritizes speech rate.
EmoInstruct-TTS uses Emotion2embed and an Instruction-Conditioned Emotion Flow Model (ICE-Flow) to generate acoustically grounded emotion representations from free-form instructions and integrate them into an LLM-based TTS pipeline.
dots.tts reports SOTA benchmark results on Seed-TTS-Eval and other tests via continuous latent-space autoregressive modeling with three listed innovations and code release.
CleanCodec reframes audio tokenization as a selective information bottleneck to encode only perceptually important features at 12.5 tokens per second, outperforming prior codecs in efficiency, speaker similarity, and intelligibility.
LaSR improves context-aware terminology recognition in speech LLMs by aligning latent CoT supervision on acoustic regions and introducing latent reasoning periods, shown on a new academic corpus to outperform standard fine-tuning without added latency.
MindVoice disentangles neural-to-speech reconstruction into semantic and acoustic pathways using pretrained priors, then fuses them with speech generation models to produce intelligible output from non-invasive recordings.
Dasheng AudioGen uses multi-view captions and a unified semantic-acoustic representation to enable end-to-end generation of mixed audio scenes from text descriptions.
A wait-think-answer controller for LALMs is trained via SFT followed by six-reward DAPO, raising row-weighted accuracy from 67.6% to 70.3% and cutting post-endpoint thinking length by 14% on synthetic spoken QA while remaining functional on real recorded audio.
Emotion embedding similarities are unsuitable for zero-shot evaluation of emotional expressiveness in speech generation due to confounding by non-emotional acoustic features.
TTS-PRISM defines a 12-dimensional perceptual schema, builds a targeted diagnostic dataset via adversarial synthesis and expert labels, and tunes an end-to-end model that outperforms generalist LLMs in human alignment on a 1,600-sample Mandarin test set while profiling six TTS paradigms.
citing papers explorer
-
WavTTS: Towards High-Quality Zero-Shot TTS via Direct Raw Waveform Modeling
WavTTS is the first raw-waveform diffusion TTS model using DiT flow matching and multi-scale mel supervision that approaches SOTA latent zero-shot performance while beating prior end-to-end models.
-
FlexiSLM: A Dynamic and Controllable Frame Rate Spoken Language Model
FlexiSLM is the first spoken language model supporting dynamic and controllable frame rates on speech input and output, outperforming fixed-rate 7B models at high quality and enabling faster inference at lower rates like 6.25 Hz.
-
JAVEDIT: Joint Audio-Visual Instruction-Guided Video Editing with Agentic Data Curation
JAVEdit-100k is the first large-scale dataset for instruction-guided joint audio-visual video editing, accompanied by JAVEditBench and the JAVEdit model that outperforms baselines on five of six metrics.
-
Unified Synthesis of Compositional Speech and Sound from Free-Form Text Prompts
PlanAudio introduces a unified autoregressive LLM framework with semantic latent chain-of-thought for generating composite speech and sound audio from free-form text, plus a new benchmark.
-
Kinetic-Optimal Scheduling with Moment Correction for Metric-Induced Discrete Flow Matching in Zero-Shot Text-to-Speech
GibbsTTS combines a training-free kinetic-optimal scheduler with finite-step moment correction in MI-DFM to deliver top naturalness and strong speaker similarity in zero-shot TTS.
-
VITA-QinYu: Expressive Spoken Language Model for Role-Playing and Singing
VITA-QinYu is the first expressive end-to-end spoken language model supporting role-playing and singing alongside conversation, trained on 15.8K hours of data and outperforming prior models on expressiveness and conversational benchmarks.
-
Talker-T2AV: Joint Talking Audio-Video Generation with Autoregressive Diffusion Modeling
Talker-T2AV achieves better lip-sync accuracy, video quality, and audio quality than dual-branch baselines by separating high-level shared autoregressive modeling from modality-specific low-level diffusion refinement in a joint audio-video generation framework.
-
MINT-Bench: A Comprehensive Multilingual Benchmark for Instruction-Following Text-to-Speech
MINT-Bench is a new benchmark using hierarchical taxonomy, multi-stage data pipeline, and hybrid evaluation to assess instruction-following TTS systems, revealing major gaps in compositional and paralinguistic controls.
-
NVBench: A Benchmark for Speech Synthesis with Non-Verbal Vocalizations
NVBench provides a standardized bilingual benchmark and evaluation protocol for assessing non-verbal vocalization generation, placement, and salience in text-to-speech systems.
-
HumDial-EIBench: A Human-Recorded Multi-Turn Emotional Intelligence Benchmark for Audio Language Models
HumDial-EIBench is a new benchmark using real human dialogues to evaluate audio language models on emotional intelligence tasks including multi-turn tracking, causal reasoning, empathy generation, and acoustic-semantic conflict resolution.
-
Knowing What to Stress: A Discourse-Conditioned Text-to-Speech Benchmark
CAST benchmark shows language models infer correct word stress from discourse context but TTS systems frequently fail to produce it in speech.
-
CapTalk: Unified Voice Design for Single-Utterance and Dialogue Speech Generation
CapTalk unifies single-utterance and dialogue voice design via utterance- and speaker-level captions plus a hierarchical variational module for stable timbre with adaptive expression.
-
UniSAE: Unified Speech Attribute Editing on Speaker, Emotion and Low-Level Content via Discrete Phonetic Posteriorgram Modelling
UniSAE unifies speaker, emotion, and multi-granularity content editing in speech via a new discrete phonetic posteriorgram representation and diffusion-based rendering.
-
HPRO: Hierarchical Progressive Reward Optimization via Preference Extraction for Emotional Text-to-Speech
HPRO uses a differentiable HD-Emo codec to extract separate content and style tokens and progressively aligns frame-, word-, and sentence-level rewards to improve emotional expressiveness in TTS while preserving intelligibility.
-
What Makes Synthetic Speech Sound Sarcastic? A Prosody-Controlled Perception Study
Controllable neural TTS reveals that loudness drives human sarcasm perception while a model prioritizes speech rate.
-
EmoInstruct-TTS: Dual-Path Instruction-Guided Emotional Speech Synthesis
EmoInstruct-TTS uses Emotion2embed and an Instruction-Conditioned Emotion Flow Model (ICE-Flow) to generate acoustically grounded emotion representations from free-form instructions and integrate them into an LLM-based TTS pipeline.
-
dots.tts Technical Report
dots.tts reports SOTA benchmark results on Seed-TTS-Eval and other tests via continuous latent-space autoregressive modeling with three listed innovations and code release.
-
CleanCodec: Efficient and Robust Speech Tokenization via Perceptually Guided Encoding
CleanCodec reframes audio tokenization as a selective information bottleneck to encode only perceptually important features at 12.5 tokens per second, outperforming prior codecs in efficiency, speaker similarity, and intelligibility.
-
LaSR: Context-Aware Speech Recognition via Latent Reasoning
LaSR improves context-aware terminology recognition in speech LLMs by aligning latent CoT supervision on acoustic regions and introducing latent reasoning periods, shown on a new academic corpus to outperform standard fine-tuning without added latency.
-
MindVoice: Reconstructing Intelligible Speech from Non-invasive Neural Signals with Pretrained Priors
MindVoice disentangles neural-to-speech reconstruction into semantic and acoustic pathways using pretrained priors, then fuses them with speech generation models to produce intelligible output from non-invasive recordings.
-
Dasheng AudioGen: A Unified Model for Generating Coherent Audio Scenes from Text
Dasheng AudioGen uses multi-view captions and a unified semantic-acoustic representation to enable end-to-end generation of mixed audio scenes from text descriptions.
-
Learning When to Think While Listening in Large Audio-Language Models
A wait-think-answer controller for LALMs is trained via SFT followed by six-reward DAPO, raising row-weighted accuracy from 67.6% to 70.3% and cutting post-endpoint thinking length by 14% on synthetic spoken QA while remaining functional on real recorded audio.
-
The False Resonance: A Critical Examination of Emotion Embedding Similarity for Speech Generation Evaluation
Emotion embedding similarities are unsuitable for zero-shot evaluation of emotional expressiveness in speech generation due to confounding by non-emotional acoustic features.
-
TTS-PRISM: A Perceptual Reasoning and Interpretable Speech Model for Fine-Grained Diagnosis
TTS-PRISM defines a 12-dimensional perceptual schema, builds a targeted diagnostic dataset via adversarial synthesis and expert labels, and tunes an end-to-end model that outperforms generalist LLMs in human alignment on a 1,600-sample Mandarin test set while profiling six TTS paradigms.
-
Audio2Tool: Speak, Call, Act -- A Dataset for Benchmarking Speech Tool Use
Audio2Tool is a new benchmark dataset that shows speech models perform well on simple commands but degrade sharply on compositional tasks and realistic acoustic noise.
-
ASPIRin: Action Space Projection for Interactivity-Optimized Reinforcement Learning in Full-Duplex Speech Language Models
ASPIRin decouples speaking timing from token content via binary action space projection and applies GRPO with rule-based rewards to optimize interactivity in SLMs without semantic collapse or repetition.
-
OmniVoice: Towards Omnilingual Zero-Shot Text-to-Speech with Diffusion Language Models
OmniVoice introduces a diffusion language model-style non-autoregressive TTS system that directly maps text to multi-codebook acoustic tokens, scaling zero-shot synthesis to over 600 languages with SOTA results on multilingual benchmarks using 581k hours of open data.
-
How to Leverage Synthetic Speech for LLM-Based ASR Systems?
Layer selection plus RIR augmentation on synthetic speech matches full real-data ASR performance using 25% real speech in SLAM-ASR.
-
ContextCodec: Content-Focused Context Guidance for Ultra-Low Bitrate Speech Coding
ContextCodec uses a dual-branch encoder with CLIP-style contrastive training on phoneme-aligned context features plus autoregressive refinement to improve quality-intelligibility at bitrates down to 500 bps.
-
End-to-End Training for Discrete Token LLM based TTS System
An end-to-end optimization framework jointly trains the speech tokenizer, LLM, FM model, and reward model for discrete-token TTS, reporting new SOTA WER of 0.78% and 1.56% on Seed-TTS-Eval with 0.6B LLM and 0.5B FM.
-
FlashTTS: Fast Streaming TTS with MTP Acceleration and X-pred Mean Flow Distillation
FlashTTS delivers a streaming TTS system using multi-track input processing and X-pred mean flow matching to reach 325 ms latency in two function evaluations while retaining zero-shot voice cloning.
-
VoxCPM2 Technical Report
VoxCPM2 scales hierarchical continuous-latent speech modeling to 2B parameters and over 2M hours of multilingual data, unifying voice cloning, style control, and continuation in one backbone with open release.
-
Do speech foundation models perceive speaker similarity as humans do?
The study compares speaker embeddings from more than 40 speech foundation models with human subjective similarity scores and identifies model factors that better align with human perception.
-
Sympatheia: Emotionally Adaptive Voice Assistant with Continuous Affect Conditioning
Sympatheia introduces a continuous affect-conditioned speech dialogue model and the Sympatheia-18k synthetic dataset, showing improved emotional appropriateness over baselines when speech cues are limited.
-
DUET: Unified Dual-Space Emotion Control for Diffusion and Flow-Matching Driven Text-to-Speech
DUET enables fine-grained emotion control in pretrained diffusion and flow-matching TTS models via unified hidden-space steering and mel-space guidance, outperforming supervised baselines on multiple backbones.
-
Raon-OpenTTS: Open Models and Data for Robust Text-to-Speech
Raon-OpenTTS provides an open 510K-hour curated speech dataset and DiT-based TTS models up to 1B parameters that achieve competitive WER and speaker similarity on benchmarks versus closed models trained on millions of hours.
-
Omni-Customizer: End-to-End MultiModal Customization for Joint Audio-Video Generation
Omni-Customizer proposes an end-to-end framework using Omni-Context Fusion, Masked TTS Cross-Attention, Semantic-Anchored Multimodal RoPE, and specialized training curricula to achieve precise multimodal identity binding in joint audio-video generation.
-
JaiTTS: A Thai Voice Cloning Model
JaiTTS-v1.0 achieves 1.94% CER on short Thai speech, beating human ground truth of 1.98%, matches humans on long speech, and wins 283 of 400 human comparisons against commercial systems.
-
MAVIN: Multi-Shot Audio-Visual Generation with Narrative Control
MAVIN proposes boundary-aware attention, ID-aware propagation, a multi-agent scripting pipeline, and the MAVINSet dataset as the first framework for multi-shot audio-visual generation with narrative control, claiming SOTA results.
-
HydraQE: OSU's Submission for the IWSLT 2026 Speech Translation Metrics Shared Task
HydraQE is a new end-to-end speech translation QE system using Qwen3-ASR backbone, sparsemax layer mixing, bidirectional Transformer, and multi-task curriculum training on human and pseudo labels that outperforms cascaded baselines.
-
PilotTTS: A Disciplined Modular Recipe for Competitive Speech Synthesis
PilotTTS achieves lowest WER 1.50% (en) and CER 0.87% (zh) plus highest speaker similarity on Seed-TTS Eval using a Q-Former conditioned autoregressive architecture and a released multi-stage open data pipeline.
-
RADAR Challenge 2026: Robust Audio Deepfake Recognition under Media Transformations
RADAR Challenge 2026 organizes a multilingual audio deepfake detection benchmark with media transformations, reporting participation from 33 development and 22 evaluation teams using EER metric.
-
EdgeFM: Efficient Edge Inference for Vision-Language Models
EdgeFM is an agent-driven VLM inference framework achieving up to 1.49x speedup over TensorRT-Edge-LLM on NVIDIA Orin and first end-to-end deployment on Horizon Journey platform.
-
One Voice, Many Tongues: Cross-Lingual Voice Cloning for Scientific Speech
Authors submit a cross-lingual voice cloning system to IWSLT 2026 using OmniVoice fine-tuned on ensemble-distilled synthetic data, reporting gains in WER, CER, and speaker similarity for scientific texts in three languages.
-
KIT's Submission to Cross-Lingual Voice Cloning in IWSLT 2026
KIT's IWSLT 2026 submission adapts a multilingual TTS model with language prompting, RL fine-tuning, and reference-conditioned lexical matching, reporting largest gains from prompting.