arXiv subjects

Mengxiao Bi

Publications and source records attributed to Mengxiao Bi.

At least 19 recordsLinked to original sources

arXiv2026

Luna-TTS Family Technical Report

Modern text-to-speech (TTS) is dominated by autoregressive (AR) codec language models, whose left-to-right decoding brings latency that grows with utterance length, error accumulation along the committed prefix, and an artificial generation order imposed on the Residual Vector Quantization (RVQ) token grid. We propose Luna-TTS Family, diffusion-language-model-based TTS systems pretrained on 1 million hours of speech across Chinese, English, Japanese, and Korean. The family is built by progressive adaptation of a pretrained AR text LLM, from causal to bidirectional and finally to block-causal attention, and comprises two variants sharing a single tokenizer, data pipeline, and 0.6B backbone lineage. Luna-TTS is fully non-autoregressive: it generates the entire RVQ token grid in a fixed number of parallel refinement steps, with zero-shot voice cloning and speech editing arising natively as infilling. Luna-TTS Realtime, derived by continual training, is autoregressive over blocks of 32 codec frames (1.28s) while denoising each block in parallel; it supports KV-cached blockwise generation and incremental audio delivery, achieving an end-to-end RTF of 0.0240 and 41.6 ms local first-block latency under the warmed serving protocol. An annealed fine-tuning stage adds explicit control over emotion and non-verbal vocalizations (NVVs), and a reinforcement-learning stage applies GRPO with policy ratios computed over the realized denoising trajectory. On Seed-TTS-Eval, Luna-TTS achieves the best results on all four metrics among compared open-source and commercial systems (0.73 CER / 79.7 SIM on test-zh, 1.49 WER / 76.8 SIM on test-en); on the harder in-the-wild CV3-Eval, it posts the lowest Mandarin and English error rates in our comparison. Against leading commercial systems, it achieves the best results on most objective, model-based, and human-rated metrics for NVV and emotion control.

cs.SD↗

arXiv2026

UrgentMOS: Unified Multi-Metric and Preference Learning for Robust Speech Quality Assessment

Automatic speech quality assessment has become increasingly important as modern speech generation systems continue to advance, while human listening tests remain costly, time-consuming, and difficult to scale. Most existing learning-based assessment models rely primarily on scarce human-annotated mean opinion score (MOS) data, which limits robustness and generalization, especially when training across heterogeneous datasets. In this work, we propose UrgentMOS, a unified speech quality assessment framework that jointly learns from diverse objective and perceptual quality metrics, while explicitly tolerating the absence of arbitrary subsets of metrics during training. By leveraging complementary quality facets under heterogeneous supervision, UrgentMOS enables effective utilization of partially annotated data and improves robustness when trained on large-scale, multi-source datasets. Beyond absolute score prediction, UrgentMOS explicitly models pairwise quality preferences by directly predicting comparative MOS (CMOS), making it well suited for preference-based evaluation scenarios commonly adopted in system benchmarking. Extensive experiments across a wide range of speech quality datasets, including simulated distortions, speech enhancement, and speech synthesis, demonstrate that UrgentMOS consistently achieves state-of-the-art performance in both absolute and comparative evaluation settings.

cs.SD↗

arXiv2026

DeepASMR: LLM-Based Zero-Shot ASMR Speech Generation for Anyone of Any Voice

While modern Text-to-Speech (TTS) systems achieve high fidelity for read-style speech, they struggle to generate Autonomous Sensory Meridian Response (ASMR), a specialized, low-intensity speech style essential for relaxation. The inherent challenges include ASMR's subtle, often unvoiced characteristics and the demand for zero-shot speaker adaptation. In this paper, we introduce DeepASMR, the first framework designed for zero-shot ASMR generation. We demonstrate that a single short snippet of a speaker's ordinary, read-style speech is sufficient to synthesize high-fidelity ASMR in their voice, eliminating the need for whispered training data from the target speaker. Methodologically, we first identify that discrete speech tokens provide a soft factorization of ASMR style from speaker timbre. Leveraging this insight, we propose a two-stage pipeline incorporating a Large Language Model (LLM) for content-style encoding and a flow-matching acoustic decoder for timbre reconstruction. Furthermore, we contribute DeepASMR-DB, a comprehensive 670-hour English-Chinese multi-speaker ASMR speech corpus, and introduce a novel evaluation protocol integrating objective metrics, human listening tests, LLM-based scoring and unvoiced speech analysis. Extensive experiments confirm that DeepASMR achieves state-of-the-art naturalness and style fidelity in ASMR generation for anyone of any voice, while maintaining competitive performance on normal speech synthesis.

cs.SD↗

arXiv2025

REF-VC: Robust, Expressive and Fast Zero-Shot Voice Conversion with Diffusion Transformers

In real-world voice conversion applications, environmental noise in source speech and user demands for expressive output pose critical challenges. Traditional ASR-based methods ensure noise robustness but suppress prosody richness, while SSL-based models improve expressiveness but suffer from timbre leakage and noise sensitivity. This paper proposes REF-VC, a noise-robust expressive voice conversion system. Key innovations include: (1) A random erasing strategy to mitigate the information redundancy inherent in SSL features, enhancing noise robustness and expressiveness; (2) Implicit alignment inspired by E2TTS to suppress non-essential feature reconstruction; (3) Integration of Shortcut Models to accelerate flow matching inference, significantly reducing to 4 steps. Experimental results demonstrate that REF-VC outperforms baselines such as Seed-VC in zero-shot scenarios on the noisy set, while also performing comparably to Seed-VC on the clean set. In addition, REF-VC can be compatible with singing voice conversion within one model.

eess.AS↗

arXiv2024

Multimodal Latent Diffusion Model for Complex Sewing Pattern Generation

Generating sewing patterns in garment design is receiving increasing attention due to its CG-friendly and flexible-editing nature. Previous sewing pattern generation methods have been able to produce exquisite clothing, but struggle to design complex garments with detailed control. To address these issues, we propose SewingLDM, a multi-modal generative model that generates sewing patterns controlled by text prompts, body shapes, and garment sketches. Initially, we extend the original vector of sewing patterns into a more comprehensive representation to cover more intricate details and then compress them into a compact latent space. To learn the sewing pattern distribution in the latent space, we design a two-step training strategy to inject the multi-modal conditions, \ie, body shapes, text prompts, and garment sketches, into a diffusion model, ensuring the generated garments are body-suited and detail-controlled. Comprehensive qualitative and quantitative experiments show the effectiveness of our proposed method, significantly surpassing previous approaches in terms of complex garment design and various body adaptability. Our project page: https://shengqiliu1.github.io/SewingLDM.

cs.CV↗

arXiv2024

Revealing Directions for Text-guided 3D Face Editing

3D face editing is a significant task in multimedia, aimed at the manipulation of 3D face models across various control signals. The success of 3D-aware GAN provides expressive 3D models learned from 2D single-view images only, encouraging researchers to discover semantic editing directions in its latent space. However, previous methods face challenges in balancing quality, efficiency, and generalization. To solve the problem, we explore the possibility of introducing the strength of diffusion model into 3D-aware GANs. In this paper, we present Face Clan, a fast and text-general approach for generating and manipulating 3D faces based on arbitrary attribute descriptions. To achieve disentangled editing, we propose to diffuse on the latent space under a pair of opposite prompts to estimate the mask indicating the region of interest on latent codes. Based on the mask, we then apply denoising to the masked latent codes to reveal the editing direction. Our method offers a precisely controllable manipulation method, allowing users to intuitively customize regions of interest with the text description. Experiments demonstrate the effectiveness and generalization of our Face Clan for various pre-trained GANs. It offers an intuitive and wide application for text-guided face editing that contributes to the landscape of multimedia content creation.

cs.CV↗

arXiv2024

E1 TTS: Simple and Fast Non-Autoregressive TTS

This paper introduces Easy One-Step Text-to-Speech (E1 TTS), an efficient non-autoregressive zero-shot text-to-speech system based on denoising diffusion pretraining and distribution matching distillation. The training of E1 TTS is straightforward; it does not require explicit monotonic alignment between the text and audio pairs. The inference of E1 TTS is efficient, requiring only one neural network evaluation for each utterance. Despite its sampling efficiency, E1 TTS achieves naturalness and speaker similarity comparable to various strong baseline models. Audio samples are available at http://e1tts.github.io/ .

eess.AS↗

arXiv2024

MacST: Multi-Accent Speech Synthesis via Text Transliteration for Accent Conversion

In accented voice conversion or accent conversion, we seek to convert the accent in speech from one another while preserving speaker identity and semantic content. In this study, we formulate a novel method for creating multi-accented speech samples, thus pairs of accented speech samples by the same speaker, through text transliteration for training accent conversion systems. We begin by generating transliterated text with Large Language Models (LLMs), which is then fed into multilingual TTS models to synthesize accented English speech. As a reference system, we built a sequence-to-sequence model on the synthetic parallel corpus for accent conversion. We validated the proposed method for both native and non-native English speakers. Subjective and objective evaluations further validate our dataset's effectiveness in accent conversion studies.

cs.SD↗

arXiv2024

HIMO: A New Benchmark for Full-Body Human Interacting with Multiple Objects

Generating human-object interactions (HOIs) is critical with the tremendous advances of digital avatars. Existing datasets are typically limited to humans interacting with a single object while neglecting the ubiquitous manipulation of multiple objects. Thus, we propose HIMO, a large-scale MoCap dataset of full-body human interacting with multiple objects, containing 3.3K 4D HOI sequences and 4.08M 3D HOI frames. We also annotate HIMO with detailed textual descriptions and temporal segments, benchmarking two novel tasks of HOI synthesis conditioned on either the whole text prompt or the segmented text prompts as fine-grained timeline control. To address these novel tasks, we propose a dual-branch conditional diffusion model with a mutual interaction module for HOI synthesis. Besides, an auto-regressive generation pipeline is also designed to obtain smooth transitions between HOI segments. Experimental results demonstrate the generalization ability to unseen object geometries and temporal compositions.

cs.CV↗

arXiv2024

DualVC 3: Leveraging Language Model Generated Pseudo Context for End-to-end Low Latency Streaming Voice Conversion

Streaming voice conversion has become increasingly popular for its potential in real-time applications. The recently proposed DualVC 2 has achieved robust and high-quality streaming voice conversion with a latency of about 180ms. Nonetheless, the recognition-synthesis framework hinders end-to-end optimization, and the instability of automatic speech recognition (ASR) model with short chunks makes it challenging to further reduce latency. To address these issues, we propose an end-to-end model, DualVC 3. With speaker-independent semantic tokens to guide the training of the content encoder, the dependency on ASR is removed and the model can operate under extremely small chunks, with cascading errors eliminated. A language model is trained on the content encoder output to produce pseudo context by iteratively predicting future frames, providing more contextual information for the decoder to improve conversion quality. Experimental results demonstrate that DualVC 3 achieves comparable performance to DualVC 2 in subjective and objective metrics, with a latency of only 50 ms.

eess.AS↗

arXiv2024

Preconditioned Nonlinear Conjugate Gradient Method for Real-time Interior-point Hyperelasticity

The linear conjugate gradient method is widely used in physical simulation, particularly for solving large-scale linear systems derived from Newton's method. The nonlinear conjugate gradient method generalizes the conjugate gradient method to nonlinear optimization, which is extensively utilized in solving practical large-scale unconstrained optimization problems. However, it is rarely discussed in physical simulation due to the requirement of multiple vector-vector dot products. Fortunately, with the advancement of GPU-parallel acceleration techniques, it is no longer a bottleneck. In this paper, we propose a Jacobi preconditioned nonlinear conjugate gradient method for elastic deformation using interior-point methods. Our method is straightforward, GPU-parallelizable, and exhibits fast convergence and robustness against large time steps. The employment of the barrier function in interior-point methods necessitates continuous collision detection per iteration to obtain a penetration-free step size, which is computationally expensive and challenging to parallelize on GPUs. To address this issue, we introduce a line search strategy that deduces an appropriate step size in a single pass, eliminating the need for additional collision detection. Furthermore, we simplify and accelerate the computations of Jacobi preconditioning and Hessian-vector product for hyperelasticity and barrier function. Our method can accurately simulate objects comprising over 100,000 tetrahedra in complex self-collision scenarios at real-time speeds.

math.OC↗

arXiv2024

EDTalk: Efficient Disentanglement for Emotional Talking Head Synthesis

Achieving disentangled control over multiple facial motions and accommodating diverse input modalities greatly enhances the application and entertainment of the talking head generation. This necessitates a deep exploration of the decoupling space for facial features, ensuring that they a) operate independently without mutual interference and b) can be preserved to share with different modal input, both aspects often neglected in existing methods. To address this gap, this paper proposes a novel Efficient Disentanglement framework for Talking head generation (EDTalk). Our framework enables individual manipulation of mouth shape, head pose, and emotional expression, conditioned on video or audio inputs. Specifically, we employ three lightweight modules to decompose the facial dynamics into three distinct latent spaces representing mouth, pose, and expression, respectively. Each space is characterized by a set of learnable bases whose linear combinations define specific motions. To ensure independence and accelerate training, we enforce orthogonality among bases and devise an efficient training strategy to allocate motion responsibilities to each space without relying on external knowledge. The learned bases are then stored in corresponding banks, enabling shared visual priors with audio input. Furthermore, considering the properties of each space, we propose an Audio-to-Motion module for audio-driven talking head synthesis. Experiments are conducted to demonstrate the effectiveness of EDTalk. We recommend watching the project website: https://tanshuai0219.github.io/EDTalk/

cs.CV↗

arXiv2024

DualVC 2: Dynamic Masked Convolution for Unified Streaming and Non-Streaming Voice Conversion

Voice conversion is becoming increasingly popular, and a growing number of application scenarios require models with streaming inference capabilities. The recently proposed DualVC attempts to achieve this objective through streaming model architecture design and intra-model knowledge distillation along with hybrid predictive coding to compensate for the lack of future information. However, DualVC encounters several problems that limit its performance. First, the autoregressive decoder has error accumulation in its nature and limits the inference speed as well. Second, the causal convolution enables streaming capability but cannot sufficiently use future information within chunks. Third, the model is unable to effectively address the noise in the unvoiced segments, lowering the sound quality. In this paper, we propose DualVC 2 to address these issues. Specifically, the model backbone is migrated to a Conformer-based architecture, empowering parallel inference. Causal convolution is replaced by non-causal convolution with dynamic chunk mask to make better use of within-chunk future information. Also, quiet attention is introduced to enhance the model's noise robustness. Experiments show that DualVC 2 outperforms DualVC and other baseline systems in both subjective and objective metrics, with only 186.4 ms latency. Our audio samples are made publicly available.

eess.AS↗

arXiv2023

Multi-GradSpeech: Towards Diffusion-based Multi-Speaker Text-to-speech Using Consistent Diffusion Models

Despite imperfect score-matching causing drift in training and sampling distributions of diffusion models, recent advances in diffusion-based acoustic models have revolutionized data-sufficient single-speaker Text-to-Speech (TTS) approaches, with Grad-TTS being a prime example. However, the sampling drift problem leads to these approaches struggling in multi-speaker scenarios in practice due to more complex target data distribution compared to single-speaker scenarios. In this paper, we present Multi-GradSpeech, a multi-speaker diffusion-based acoustic models which introduces the Consistent Diffusion Model (CDM) as a generative modeling approach. We enforce the consistency property of CDM during the training process to alleviate the sampling drift problem in the inference stage, resulting in significant improvements in multi-speaker TTS performance. Our experimental results corroborate that our proposed approach can improve the performance of different speakers involved in multi-speaker TTS compared to Grad-TTS, even outperforming the fine-tuning approach. Audio samples are available at https://welkinyang.github.io/multi-gradspeech/

eess.AS↗

arXiv2023

DualVC: Dual-mode Voice Conversion using Intra-model Knowledge Distillation and Hybrid Predictive Coding

Voice conversion is an increasingly popular technology, and the growing number of real-time applications requires models with streaming conversion capabilities. Unlike typical (non-streaming) voice conversion, which can leverage the entire utterance as full context, streaming voice conversion faces significant challenges due to the missing future information, resulting in degraded intelligibility, speaker similarity, and sound quality. To address this challenge, we propose DualVC, a dual-mode neural voice conversion approach that supports both streaming and non-streaming modes using jointly trained separate network parameters. Furthermore, we propose intra-model knowledge distillation and hybrid predictive coding (HPC) to enhance the performance of streaming conversion. Additionally, we incorporate data augmentation to train a noise-robust autoregressive decoder, improving the model's performance on long-form speech conversion. Experimental results demonstrate that the proposed model outperforms the baseline models in the context of streaming voice conversion, while maintaining comparable performance to the non-streaming topline system that leverages the complete context, albeit with a latency of only 252.8 ms.

eess.AS↗

arXiv2022

Expressive-VC: Highly Expressive Voice Conversion with Attention Fusion of Bottleneck and Perturbation Features

Voice conversion for highly expressive speech is challenging. Current approaches struggle with the balancing between speaker similarity, intelligibility and expressiveness. To address this problem, we propose Expressive-VC, a novel end-to-end voice conversion framework that leverages advantages from both neural bottleneck feature (BNF) approach and information perturbation approach. Specifically, we use a BNF encoder and a Perturbed-Wav encoder to form a content extractor to learn linguistic and para-linguistic features respectively, where BNFs come from a robust pre-trained ASR model and the perturbed wave becomes speaker-irrelevant after signal perturbation. We further fuse the linguistic and para-linguistic features through an attention mechanism, where speaker-dependent prosody features are adopted as the attention query, which result from a prosody encoder with target speaker embedding and normalized pitch and energy of source speech as input. Finally the decoder consumes the integrated features and the speaker-dependent prosody feature to generate the converted speech. Experiments demonstrate that Expressive-VC is superior to several state-of-the-art systems, achieving both high expressiveness captured from the source speech and high speaker similarity with the target speaker; meanwhile intelligibility is well maintained.

eess.AS↗

arXiv2022

Learn2Sing 2.0: Diffusion and Mutual Information-Based Target Speaker SVS by Learning from Singing Teacher

Building a high-quality singing corpus for a person who is not good at singing is non-trivial, thus making it challenging to create a singing voice synthesizer for this person. Learn2Sing is dedicated to synthesizing the singing voice of a speaker without his or her singing data by learning from data recorded by others, i.e., the singing teacher. Inspired by the fact that pitch is the key style factor to distinguish singing from speaking voice, the proposed Learn2Sing 2.0 first generates the preliminary acoustic feature with averaged pitch value in the phone level, which allows the training of this process for different styles, i.e., speaking or singing, share same conditions except for the speaker information. Then, conditioned on the specific style, a diffusion decoder, which is accelerated by a fast sampling algorithm during the inference stage, is adopted to gradually restore the final acoustic feature. During the training, to avoid the information confusion of the speaker embedding and the style embedding, mutual information is employed to restrain the learning of speaker embedding and style embedding. Experiments show that the proposed approach is capable of synthesizing high-quality singing voice for the target speaker without singing data with 10 decoding steps.

cs.SD↗

arXiv2022

VISinger: Variational Inference with Adversarial Learning for End-to-End Singing Voice Synthesis

In this paper, we propose VISinger, a complete end-to-end high-quality singing voice synthesis (SVS) system that directly generates audio waveform from lyrics and musical score. Our approach is inspired by VITS, which adopts VAE-based posterior encoder augmented with normalizing flow-based prior encoder and adversarial decoder to realize complete end-to-end speech generation. VISinger follows the main architecture of VITS, but makes substantial improvements to the prior encoder based on the characteristics of singing. First, instead of using phoneme-level mean and variance of acoustic features, we introduce a length regulator and a frame prior network to get the frame-level mean and variance on acoustic features, modeling the rich acoustic variation in singing. Second, we further introduce an F0 predictor to guide the frame prior network, leading to stabler singing performance. Finally, to improve the singing rhythm, we modify the duration predictor to specifically predict the phoneme to note duration ratio, helped with singing note normalization. Experiments on a professional Mandarin singing corpus show that VISinger significantly outperforms FastSpeech+Neural-Vocoder two-stage approach and the oracle VITS; ablation study demonstrates the effectiveness of different contributions.

eess.AS↗