6 citations · 8 across the 11 of their papers we have counts for
11 papers
Breaking the Quality--Intelligibility Trade-off in Streaming Target Speaker Extraction via Deep-Feature-Anchored Preference Optimization
Shuhai Peng, Jinjiang Liu, Hui Lu +5
Generative streaming models for Target Speaker Extraction (TSE) commonly exhibit a quality--intelligibility trade-off, wherein naive optimization for perceptual audio quality tends…
SemaVoice: Semantic-Aware Continuous Autoregressive Speech Synthesis
Huimeng Wang, Hui Lu, Jiajun Deng +7
Continuous autoregressive speech synthesis has recently emerged as a promising direction for zero-shot text-to-speech (TTS). However, existing methods still suffer from a fundament…
How Should LLMs Listen While Speaking? A Study of User-Stream Routing in Full-Duplex Spoken Dialogue
Hui Lu, Xueyuan Chen, Huimeng Wang +4
Full-duplex spoken dialogue requires a model to keep listening while generating its own spoken response. This is challenging for large language models (LLMs), which are designed to…
StarTSE: Towards Streaming Target Speaker Extraction via Chunk-wise Interleaved Splicing of Autoregressive Language Model
Shuhai Peng, Hui Lu, Jinjiang Liu +8
While generative models have set new benchmarks for Target Speaker Extraction (TSE), their inherent reliance on global context precludes deployment in real-time applications. Direc…
Addressing Index Collapse of Large-Codebook Speech Tokenizer with Dual-Decoding Product-Quantized Variational Auto-Encoder
Haohan Guo, Fenglong Xie, Dongchao Yang +3
VQ-VAE, as a mainstream approach of speech tokenizer, has been troubled by ``index collapse'', where only a small number of codewords are activated in large codebooks. This work pr…
Towards High-Quality Neural TTS for Low-Resource Languages by Learning Compact Speech Representations
Haohan Guo, Fenglong Xie, Xixin Wu +2
This paper aims to enhance low-resource TTS by reducing training data requirements using compact speech representations. A Multi-Stage Multi-Codebook (MSMC) VQ-GAN is trained to le…
Disentangled Speech Representation Learning for One-Shot Cross-lingual Voice Conversion Using -VAE
Hui Lu, Disong Wang, Xixin Wu +3
We propose an unsupervised learning method to disentangle speech into content representation and speaker identity representation. We apply this method to the challenging one-shot c…
A Multi-Scale Time-Frequency Spectrogram Discriminator for GAN-based Non-Autoregressive TTS
Haohan Guo, Hui Lu, Xixin Wu +1
The generative adversarial network (GAN) has shown its outstanding capability in improving Non-Autoregressive TTS (NAR-TTS) by adversarially training it with an extra model that di…
Speaker Identity Preservation in Dysarthric Speech Reconstruction by Adversarial Speaker Adaptation
Disong Wang, Songxiang Liu, Xixin Wu +4
Dysarthric speech reconstruction (DSR), which aims to improve the quality of dysarthric speech, remains a challenge, not only because we need to restore the speech to be normal, bu…
Channel-wise Gated Res2Net: Towards Robust Detection of Synthetic Speech Attacks
Xu Li, Xixin Wu, Hui Lu +2
Existing approaches for anti-spoofing in automatic speaker verification (ASV) still lack generalizability to unseen attacks. The Res2Net approach designs a residual-like connection…
VAENAR-TTS: Variational Auto-Encoder based Non-AutoRegressive Text-to-Speech Synthesis
Hui Lu, Zhiyong Wu, Xixin Wu +4
This paper describes a variational auto-encoder based non-autoregressive text-to-speech (VAENAR-TTS) model. The autoregressive TTS (AR-TTS) models based on the sequence-to-sequence…