Recently, zero-shot text-to-speech (TTS) has enabled high-fidelity and expressive speech synthesis, but it often fails to imitate unseen speaking styles from uncommon scenarios (e.g., crosstalk, dialects). Moreover, fine-tuning pretrained models requires large, high-quality datasets, limiting rapid personalization. We propose VoiceTTA, a reinforcement learning-based test-time adaptation (TTA) method that improves voice imitation of pretrained zero-shot TTS models. VoiceTTA introduces two style rewards based on coefficient-of-variation differences of F0 and energy, combined with speaker similarity and intelligibility (WER from a pretrained Whisper model), and optimizes learnable prefixes via group relative preference optimization (GRPO) in a flow matching-based model at inference time. Extensive experiments demonstrate substantial improvements on uncommon speech prompts, outperforming state-of-the-art baselines. Audio samples are available at https://voicetta.pages.dev/
StyleSpeech: Self-supervised Style Enhancing with VQ-VAE-based Pre-training for Expressive Audiobook Speech Synthesis
The Chinese University of Hong Kong（香港中文大学）；Tsinghua-CUHK Joint Research Center for Media Sciences, Technologies and Systems（清华-香港中文大学媒体科学、技术与系统联合研究中心）；Tsinghua University（清华大学）；Microsoft（微软）
ECTSpeech: Enhancing Efficient Speech Synthesis via Easy Consistency Tuning
School of Computer Science and Technology, Xinjiang University（新疆大学计算机科学与技术学院）；Joint International Research Laboratory of Silk Road Multilingual Cognitive Computing（丝绸之路多语种认知计算联合国际实验室）；Tsinghua University（清华大学）；Tianjin University of Technology（天津工业大学）
Exploiting rich linguistic information in raw text is crucial for expressive text-to-speech (TTS). As large scale pre-trained text representation develops, bidirectional encoder representations from Transformers (BERT) has been proven to embody semantic information and employed to TTS recently. However, original or simply fine-tuned BERT embeddings still cannot provide sufficient semantic knowledge that expressive TTS models should take into account. In this paper, we propose a word-level semantic representation enhancing method based on dependency structure and pre-trained BERT embedding. The…
University of Crete（克里特大学）；Foundation for Research and Technology - Hellas（希腊研究与技术基金会）
The increased adoption of digital assistants makes text-to-speech (TTS) synthesis systems an indispensable feature of modern mobile devices. It is hence desirable to build a system capable of generating highly intelligible speech in the presence of noise. Past studies have investigated style conversion in TTS synthesis, yet degraded synthesized quality often leads to worse intelligibility. To overcome such limitations, we proposed a novel transfer learning approach using Tacotron and WaveRNN based TTS synthesis. The proposed speech system exploits two modification strategies: (a) Lombard speak…
National Institute of Informatics（国立信息学研究所）；SOKENDAI（综合研究大学院大学）；University of Science and Technology of China（中国科学技术大学）
High-quality speech corpora are essential foundations for most speech applications. However, such speech data are expensive and limited since they are collected in professional recording environments. In this work, we propose an encoder-decoder neural network to automatically enhance low-quality recordings to professional high-quality recordings. To address channel variability, we first filter out the channel characteristics from the original input audio using the encoder network with adversarial training. Next, we disentangle the channel factor from a reference audio. Conditioned on this fact…
Expressive text-to-speech (TTS) has become a hot research topic recently, mainly focusing on modeling prosody in speech. Prosody modeling has several challenges: 1) the extracted pitch used in previous prosody modeling works have inevitable errors, which hurts the prosody modeling; 2) different attributes of prosody (e.g., pitch, duration and energy) are dependent on each other and produce the natural prosody together; and 3) due to high variability of prosody and the limited amount of high-quality data for TTS training, the distribution of prosody cannot be fully shaped. To tackle these issue…
Enhancing Crowdsourced Audio for Text-to-Speech Models
Barcelona Supercomputing Center (BSC)（巴塞罗那超级计算中心）；Centre de Llenguatge i Computació (CLiC), Universitat de Barcelona（巴塞罗那大学语言与计算中心）；Universität Zürich（苏黎世大学）
本文针对众包音频质量不足的问题，在 Common Voice 加泰罗尼亚语子集上提出音频增强加 NISQA 自动过滤的去噪流水线，使扩散 TTS 模型的 UTMOS 分数提升 0.4。
Text-to-Speech synthesis systems are generally evaluated using Mean Opinion Score (MOS) tests, where listeners score samples of synthetic speech on a Likert scale. A major drawback of MOS tests is that they only offer a general measure of overall quality-i.e., the naturalness of an utterance-and so cannot tell us where exactly synthesis errors occur. This can make evaluation of the appropriateness of prosodic variation within utterances inconclusive. To address this, we propose a novel evaluation method based on the Rapid Prosody Transcription paradigm. This allows listeners to mark the locati…
Tsinghua-CUHK Joint Research Center for Media Sciences, Technologies and Systems（清华-中大媒体科学、技术与系统联合研究中心）；Shenzhen International Graduate School, Tsinghua University（清华大学深圳国际研究生院）；The Chinese University of Hong Kong（香港中文大学）；AI Lab, Tencent（腾讯人工智能实验室）
Comparing with traditional text-to-speech (TTS) systems, conversational TTS systems are required to synthesize speeches with proper speaking style confirming to the conversational context. However, state-of-the-art context modeling methods in conversational TTS only model the textual information in context with a recurrent neural network (RNN). Such methods have limited ability in modeling the inter-speaker influence in conversations, and also neglect the speaking styles and the intra-speaker inertia inside each speaker. Inspired by DialogueGCN and its superiority in modeling such conversation…
Koç University（科奇大学）；College of Engineering, Koç University（科奇大学工程学院）
We present a novel method for training a social robot to generate backchannels during human-robot interaction. We address the problem within an off-policy reinforcement learning framework, and show how a robot may learn to produce non-verbal backchannels like laughs, when trained to maximize the engagement and attention of the user. A major contribution of this work is the formulation of the problem as a Markov decision process (MDP) with states defined by the speech activity of the user and rewards generated by quantified engagement levels. The problem that we address falls into the class of …
Neural TTS has shown it can generate high quality synthesized speech. In this paper, we investigate the multi-speaker latent space to improve neural TTS for adapting the system to new speakers with only several minutes of speech or enhancing a premium voice by utilizing the data from other speakers for richer contextual coverage and better generalization. A multi-speaker neural TTS model is built with the embedded speaker information in both spectral and speaker latent space. The experimental results show that, with less than 5 minutes of training data from a new speaker, the new model can ach…
In this paper, we propose a deep learning (DL)-based parameter enhancement method for a mixed excitation linear prediction (MELP) speech codec in noisy communication environment. Unlike conventional speech enhancement modules that are designed to obtain clean speech signal by removing noise components before speech codec processing, the proposed method directly enhances codec parameters on either the encoder or decoder side. As the proposed method has been implemented by a small network without any additional processes required in conventional enhancement systems, e.g., time-frequency (T-F) an…
Delft University of Technology（代尔夫特理工大学）；Netherlands Cancer Institute（荷兰癌症研究所）；University of Amsterdam（阿姆斯特丹大学）
In this paper, we investigate several existing and a new state-of-the-art generative adversarial network-based (GAN) voice conversion method for enhancing dysarthric speech for improved dysarthric speech recognition. We compare key components of existing methods as part of a rigorous ablation study to find the most effective solution to improve dysarthric speech recognition. We find that straightforward signal processing methods such as stationary noise removal and vocoder-based time stretching lead to dysarthric speech recognition results comparable to those obtained when using state-of-the-a…
Unlocking Foundation Models for Privacy-Enhancing Speech Understanding: An Early Study on Low Resource Speech Training Leveraging Label-guided Synthetic Speech Content
Beyond saliency: enhancing explanation of speech emotion recognition with expert-referenced acoustic cues
Center for Cognitive Interaction Technology (CITEC), Bielefeld University, Germany（认知交互技术中心（CITEC），比勒菲尔德大学，德国）；Cognitive Systems Lab, University of Bremen, Germany（认知系统实验室，不莱梅大学，德国）
Density Adaptive Attention-based Speech Network: Enhancing Feature Understanding for Mental Health Disorders
James Silberrad Brown Center for Artificial Intelligence（詹姆斯·西尔伯拉德·布朗人工智能中心）；Carnegie Mellon University（卡内基梅隆大学）；Stanford University（斯坦福大学）；Amazon GenAI（亚马逊生成式人工智能）
We enhance the vanilla adversarial training method for unsupervised Automatic Speech Recognition (ASR) by a diffusion-GAN. Our model (1) injects instance noises of various intensities to the generator's output and unlabeled reference text which are sampled from pretrained phoneme language models with a length constraint, (2) asks diffusion timestep-dependent discriminators to separate them, and (3) back-propagates the gradients to update the generator. Word/phoneme error rate comparisons with wav2vec-U under Librispeech (3.1% for test-clean and 5.6% for test-other), TIMIT and MLS datasets, sho…
Enhancing Indonesian Automatic Speech Recognition: Evaluating Multilingual Models with Diverse Speech Variabilities
Institut Teknologi Bandung（万隆理工学院）；University of Indonesia（印度尼西亚大学）；Japan Advanced Institute of Science and Technology（日本先进科学技术学院）；Nara Institute of Science and Technology（奈良科学技术大学）
