arXiv subjects

Huiyan Li

Publications and source records attributed to Huiyan Li.

7 recordsLinked to original sources

arXiv2026

StellarTTS: Sparse Temporal Embedding for Low-Latency and Robust Speech Synthesis

The trade-off between robustness, latency, and prosody critically challenges text-to-speech (TTS) systems. Autoregressive models, despite fidelity, are slow and error-prone; non-autoregressive (NAR) alternatives, while fast, often sacrifice prosodic naturalness via rigid alignments. This paper introduces StellarTTS, a novel mobile-optimized NAR TTS framework based on a sparse temporal embedding strategy, enabling granular control of phoneme duration, pronunciation, and prosody. Furthermore, we propose a semantic-aware codec that facilitates efficient single-stage decoding. Conditioned on the sparse temporal embedding, our 83M-parameter lightweight masked generative transformer achieves a real-time factor (RTF) of 0.08. Experiments demonstrate that StellarTTS attains lower latency and stronger robustness compared to state-of-the-art TTS systems, while maintaining competitive performance in audio quality, prosodic naturalness, and speaker similarity.

cs.SD↗

arXiv2025

Novel Parasitic Dual-Scale Modeling for Efficient and Accurate Multilingual Speech Translation

Recent advancements in speech-to-text translation have led to the development of multilingual models capable of handling multiple language pairs simultaneously. However, these unified models often suffer from large parameter sizes, making it challenging to balance inference efficiency and performance, particularly in local deployment scenarios. We propose an innovative Parasitic Dual-Scale Approach, which combines an enhanced speculative sampling method with model compression and knowledge distillation techniques. Building on the Whisper Medium model, we enhance it for multilingual speech translation into whisperM2M, and integrate our novel KVSPN module, achieving state-of-the-art (SOTA) performance across six popular languages with improved inference efficiency. KVSPN enables a 40\% speedup with no BLEU score degradation. Combined with distillation methods, it represents a 2.6$\times$ speedup over the original Whisper Medium with superior performance.

cs.CL↗

arXiv2025

MFLA: Monotonic Finite Look-ahead Attention for Streaming Speech Recognition

Applying large pre-trained speech models like Whisper has shown promise in reducing training costs for various speech tasks. However, integrating these models into streaming systems remains a challenge. This paper presents a novel prefix-to-prefix training framework for streaming recognition by fine-tuning the Whisper. We introduce the Continuous Integrate-and-Fire mechanism to establish a quasi-monotonic alignment between continuous speech sequences and discrete text tokens. Additionally, we design Monotonic Finite Look-ahead Attention, allowing each token to attend to infinite left-context and finite right-context from the speech sequences. We also employ the wait-k decoding strategy to simplify the decoding process while ensuring consistency between training and testing. Our theoretical analysis and experiments demonstrate that this approach achieves a controllable trade-off between latency and quality, making it suitable for various streaming applications.

cs.CL↗

arXiv2021

Unsupervised cross-user adaptation in taste sensation recognition based on surface electromyography with conformal prediction and domain regularized component analysis

Human taste sensation can be qualitatively described with surface electromyography. However, the pattern recognition models trained on one subject (the source domain) do not generalize well on other subjects (the target domain). To improve the generalizability and transferability of taste sensation models developed with sEMG data, two methods were innovatively applied in this study: domain regularized component analysis (DRCA) and conformal prediction with shrunken centroids (CPSC). The effectiveness of these two methods was investigated independently in an unlabeled data augmentation process with the unlabeled data from the target domain, and the same cross-user adaptation pipeline were conducted on six subjects. The results show that DRCA improved the classification accuracy on six subjects (p < 0.05), compared with the baseline models trained only with the source domain data;, while CPSC did not guarantee the accuracy improvement. Furthermore, the combination of DRCA and CPSC presented statistically significant improvement (p < 0.05) in classification accuracy on six subjects. The proposed strategy combining DRCA and CPSC showed its effectiveness in addressing the cross-user data distribution drift in sEMG-based taste sensation recognition application. It also shows the potential in more cross-user adaptation applications.

q-bio.QM↗

arXiv2021

Sequence-to-Sequence Voice Reconstruction for Silent Speech in a Tonal Language

Silent Speech Decoding (SSD), based on articulatory neuromuscular activities, has become a prevalent task of Brain-Computer Interface (BCI) in recent years. Many works have been devoted to decoding surface electromyography (sEMG) from articulatory neuromuscular activities. However, restoring silent speech in tonal languages such as Mandarin Chinese is still difficult. This paper proposes an optimized Sequence-to-Sequence (Seq2Seq) approach to synthesize voice from the sEMG-based silent speech. We extract duration information to regulate the sEMG-based silent speech using the audio length. Then, we provide a deep-learning model with an encoder-decoder structure and a state-of-art vocoder to generate the audio waveform. Experiments based on six Mandarin Chinese speakers demonstrate that the proposed model can successfully decode silent speech in Mandarin Chinese and achieve a character error rate (CER) of 6.41% on average with human evaluation.

cs.SD↗

arXiv2015

Inducing chimera-synchronization in symmetric complex networks

In a recent study of chaos synchronization in symmetric complex networks [Pecora \textit{et al}., Nat. Commun. {\bf 5}, 4079 (2014)], it is found that stable synchronous clusters may coexist with many non-synchronous nodes in the asynchronous regime, resembling the chimera state observed in regular networks of non-locally coupled periodic oscillators. Although of practical significance, this new type of state, namely the chimera-synchronization state, is hardly generated for the general complex networks, due to either the topological instabilities or the weak coupling strength. Here, based on the strategy of pinning coupling, we propose an effective method for inducing chimera-synchronization in symmetric complex network of coupled chaotic oscillators. We are able to argue mathematically that, by pinning a group of nodes satisfying permutation symmetry, there always exits a critical pinning strength beyond which the unstable chimera-synchronization states can be successfully induced. The feasibility and efficiency of the control method are verified by numerical simulations of both artificial and real-world complex networks, with the numerical results well fitted by the theoretical predictions.

nlin.CD↗

arXiv2014

Desynchronization in an ensemble of globally coupled chaotic bursting neuronal oscillators by dynamic delayed feedback control

In this paper, we propose a dynamic delayed feedback control approach for desynchronization of chaotic-bursting synchronous activities in an ensemble of globally coupled neuronal oscillators. We demonstrate that the difference signal between an ensemble's mean field and its time delayed state, filtered and fed back to the ensemble, can suppress the self-synchronization in the ensemble. These individual units are decoupled and stabilized at the desired desynchronized states while the stimulation signal reduces to the noise level. The effectiveness of the method is illustrated by examples of two different populations of globally coupled chaotic-bursting neurons. The proposed method has potential for mild, effective and demand-controlled therapy of neurological diseases characterized by pathological synchronization.

nlin.CD↗