PRML Lab. Speech Team

FillerSpeech: Towards Human-Like Text-to-Speech Synthesis with Filler Injection and Filler Style Control

HierSpeech++: Bridging the Gap between Semantic and Acoustic Representation by Hierarchical Variational Inference for Zero-shot Speech Synthesis

Spotlight-TTS: Spotlighting the Style via Voiced-Aware Style Extraction and Style Direction Adjustment for Expressive Text-to-Speech

EmoSphere-SER: Enhancing Speech Emotion Recognition through Spherical Representation with Auxiliary Classification

DiEmo-TTS: Disentangled Emotion Representations via Self-Supervised Distillation for Cross-Speaker Emotion Transfer in Text-to-Speech

VibE-SVC: Vibrato Extraction with High-frequency F0 Contour for Singing Voice Conversion

Hierarchical Diffusion Model for Zero-Shot Singing Voice Synthesis with MIDI Priors

FLowHigh: Towards Efficient and High-Quality Audio Super-Resolution with Single-Step Flow Matching

JELLY: Joint Emotion Recognition and Context Reasoning with LLMs for Conversational Speech Synthesis

DurFlex-EVC: Duration-Flexible Emotional Voice Conversion Leveraging Discrete Representations Without Text Alignment

UnitCorrect: Unit-based Mispronunciation Correcting System with a DTW-based Detection

EmoSphere++: Emotion-Controllable Zero-Shot Text-to-Speech via Emotion-Adaptive Spherical Vector

PeriodWave: Multi-Period Flow Matching for High-Fidelity Waveform Generation

PromotiCon: Prompt-based Emotion Controllable Text-to-Speech via Prompt Generation and Matching

EmoSphere-TTS: Emotional Style and Intensity Modeling via Spherical Emotion Vector for Controllable Emotional Text-to-Speech

DiffProsody: Diffusion-Based Latent Prosody Generation for Expressive Speech Synthesis With Prosody Conditional Adversarial Training

Audio Super-Resolution With Robust Speech Representation Learning of Masked Autoencoder

TranSentence: Speech-to-Speech Translation via Language-agnostic Sentence-level Speech Encoding without Language-parallel Data

MIDI-Voice: Expressive Zero-shot Singing Voice Synthesis via MIDI-driven Priors

DDDM-VC: Decoupled Denoising Diffusion Models with Disentangled Representation and Prior Mixup for Verified Robust Voice Conversion

HierVST: Hierarchical Adaptive Zero-shot Voice Style Transfer

Diff-HierVC: Diffusion-based Hierarchical Voice Conversion with Robust Pitch Generation and Masked Prior for Zero-shot Speaker Adaptation

PauseSpeech: Natural Speech Synthesis via Pre-trained Language Model and Pause-based Prosody Modeling

HierSpeech: Bridging the Gap between Text and Speech by Hierarchical Variational Inference using Self-supervised Representation for Speech Synthesis

Duration Controllable Voice Conversion via Phoneme-Based Information Bottleneck

StyleVC: Non-parallel Voice Conversion with Adversarial Style Generalization

EmoQ-TTS: Emotion intensity Quantization for Fine-grained Controllable Emotional Text-to-Speech

Fre-GAN 2: Fast and Efficient Frequency-consistent Audio Synthesis

PVAE-TTS: High-Quality Adaptive Text-to-Speech via Progressive Variational Autoencoder

VoiceMixer: Adversarial Voice Style Mixup

Multi-SpectroGAN: High-Diversity and High-Fidelity Spectrogram Generation with Adversarial Style Recombination for Speech Synthesis

GC-TTS: Few-shot Speaker Adaptation with Geometric Constraints

Reinforce-Aligner: Reinforcement Alignment Search for Robust End-to-End Text-to-Speech

Fre-GAN: Adversarial Frequency-consistent Audio Synthesis

Audio dequantization for high fidelity audio generation in flow-based neural vocoder