arXiv subjects

Zhiyuan Tang

Publications and source records attributed to Zhiyuan Tang.

At least 19 recordsLinked to original sources

arXiv2026

Chain of Correction for Full-text Speech Recognition with Large Language Models

Full-text error correction with Large Language Models (LLMs) for Automatic Speech Recognition (ASR) is attracting increased attention for its ability to address a wide range of error types, such as punctuation restoration and inverse text normalization, across long context. However, challenges remain regarding stability, controllability, completeness, and fluency. To mitigate these issues, this paper proposes the Chain of Correction (CoC), which uses a multi-turn chat format to correct errors segment by segment, guided by pre-recognized text and full-text context for better semantic understanding. Utilizing the open-sourced ChFT dataset, we fine-tune a pre-trained LLM to evaluate CoC's performance. Experiments show that CoC significantly outperforms baseline and benchmark systems in correcting full-text ASR outputs. We also analyze correction thresholds to balance under-correction and over-rephrasing, extrapolate CoC on extra-long ASR outputs, and explore using other types of information to guide error correction.

cs.CL↗

arXiv2026

Match Made with Matrix Completion: Efficient Learning under Matching Interference

Matching markets face increasing needs to learn the matching qualities between demand and supply for effective design of matching policies. In practice, the matching rewards are high-dimensional due to the growing diversity of participants. We leverage a natural low-rank matrix structure of the matching rewards in these two-sided markets, and propose to utilize matrix completion to accelerate reward learning with limited offline data. A unique property for matrix completion in this setting is that the entries of the reward matrix are observed with matching interference -- i.e., the entries are not observed independently but dependently due to matching or budget constraints. Such matching dependence renders unique technical challenges, such as sub-optimality or inapplicability of the existing analytical tools in the matrix completion literature, since they typically rely on sample independence. In this paper, we first show that standard nuclear norm regularization remains theoretically effective under matching interference. We provide a near-optimal Frobenius norm guarantee in this setting, coupled with a new analytical technique. Next, to guide certain matching decisions, we develop a novel ``double-enhanced'' estimator, based off the nuclear norm estimator, with a near-optimal entry-wise guarantee. Our double-enhancement procedure can apply to broader sampling schemes even with dependence, which may be of independent interest. Additionally, we extend our approach to online learning settings with matching constraints such as optimal matching and stable matching, and present improved regret bounds in matrix dimensions. Finally, we demonstrate the practical value of our methods using both synthetic data and real data of labor markets.

stat.ML↗

arXiv2025

Cross-Modal Knowledge Distillation for Speech Large Language Models

In this work, we present the first systematic evaluation of catastrophic forgetting and modality inequivalence in speech large language models, showing that introducing speech capabilities can degrade knowledge and reasoning even when inputs remain textual, and performance further decreases with spoken queries. To address these challenges, we propose a cross-modal knowledge distillation framework that leverages both text-to-text and speech-to-text channels to transfer knowledge from a text-based teacher model to a speech LLM. Extensive experiments on dialogue and audio understanding tasks validate the effectiveness of our approach in preserving textual knowledge, improving cross-modal alignment, and enhancing reasoning in speech-based interactions.

cs.CL↗

arXiv2024

Full-text Error Correction for Chinese Speech Recognition with Large Language Model

Large Language Models (LLMs) have demonstrated substantial potential for error correction in Automatic Speech Recognition (ASR). However, most research focuses on utterances from short-duration speech recordings, which are the predominant form of speech data for supervised ASR training. This paper investigates the effectiveness of LLMs for error correction in full-text generated by ASR systems from longer speech recordings, such as transcripts from podcasts, news broadcasts, and meetings. First, we develop a Chinese dataset for full-text error correction, named ChFT, utilizing a pipeline that involves text-to-speech synthesis, ASR, and error-correction pair extractor. This dataset enables us to correct errors across contexts, including both full-text and segment, and to address a broader range of error types, such as punctuation restoration and inverse text normalization, thus making the correction process comprehensive. Second, we fine-tune a pre-trained LLM on the constructed dataset using a diverse set of prompts and target formats, and evaluate its performance on full-text error correction. Specifically, we design prompts based on full-text and segment, considering various output formats, such as directly corrected text and JSON-based error-correction pairs. Through various test settings, including homogeneous, up-to-date, and hard test sets, we find that the fine-tuned LLMs perform well in the full-text setting with different prompts, each presenting its own strengths and weaknesses. This establishes a promising baseline for further research. The dataset is available on the website.

cs.CL↗

arXiv2024

MaLa-ASR: Multimedia-Assisted LLM-Based ASR

As more and more information-rich data like video become available, utilizing multi-modal auxiliary information to enhance audio tasks has sparked widespread research interest. The recent surge in research on LLM-based audio models provides fresh perspectives for tackling audio tasks. Given that LLM can flexibly ingest multiple inputs, we propose MaLa-ASR, an LLM-based ASR model that can integrate textual keywords extracted from presentation slides to improve recognition of conference content. MaLa-ASR yields average WERs of 9.4% and 11.7% on the L95 and S95 subsets of the SlideSpeech corpus, representing a significant relative WER drop of 27.9% and 44.7% over the baseline model reported in SlideSpeech. MaLa-ASR underscores LLM's strong performance in speech tasks and the capability to integrate auxiliary information conveniently. By adding keywords to the input prompt, the biased word error rate (B-WER) reduces relatively by 46.0% and 44.2%, establishing a new SOTA on this dataset.

eess.AS↗

arXiv2021

Semantic Data Augmentation for End-to-End Mandarin Speech Recognition

End-to-end models have gradually become the preferred option for automatic speech recognition (ASR) applications. During the training of end-to-end ASR, data augmentation is a quite effective technique for regularizing the neural networks. This paper proposes a novel data augmentation technique based on semantic transposition of the transcriptions via syntax rules for end-to-end Mandarin ASR. Specifically, we first segment the transcriptions based on part-of-speech tags. Then transposition strategies, such as placing the object in front of the subject or swapping the subject and the object, are applied on the segmented sentences. Finally, the acoustic features corresponding to the transposed transcription are reassembled based on the audio-to-text forced-alignment produced by a pre-trained ASR system. The combination of original data and augmented one is used for training a new ASR system. The experiments are conducted on the Transformer[2] and Conformer[3] based ASR. The results show that the proposed method can give consistent performance gain to the system. Augmentation related issues, such as comparison of different strategies and ratios for data combination are also investigated.

eess.AS↗

arXiv2020

Can We Trust Deep Speech Prior?

Recently, speech enhancement (SE) based on deep speech prior has attracted much attention, such as the variational auto-encoder with non-negative matrix factorization (VAE-NMF) architecture. Compared to conventional approaches that represent clean speech by shallow models such as Gaussians with a low-rank covariance, the new approach employs deep generative models to represent the clean speech, which often provides a better prior. Despite the clear advantage in theory, we argue that deep priors must be used with much caution, since the likelihood produced by a deep generative model does not always coincide with the speech quality. We designed a comprehensive study on this issue and demonstrated that based on deep speech priors, a reasonable SE performance can be achieved, but the results might be suboptimal. A careful analysis showed that this problem is deeply rooted in the disharmony between the flexibility of deep generative models and the nature of the maximum-likelihood (ML) training.

cs.SD↗

arXiv2020

AP20-OLR Challenge: Three Tasks and Their Baselines

This paper introduces the fifth oriental language recognition (OLR) challenge AP20-OLR, which intends to improve the performance of language recognition systems, along with APSIPA Annual Summit and Conference (APSIPA ASC). The data profile, three tasks, the corresponding baselines, and the evaluation principles are introduced in this paper. The AP20-OLR challenge includes more languages, dialects and real-life data provided by Speechocean and the NSFC M2ASR project, and all the data is free for participants. The challenge this year still focuses on practical and challenging problems, with three tasks: (1) cross-channel LID, (2) dialect identification and (3) noisy LID. Based on Kaldi and Pytorch, recipes for i-vector and x-vector systems are also conducted as baselines for the three tasks. These recipes will be online-published, and available for participants to configure LID systems. The baseline results on the three tasks demonstrate that those tasks in this challenge are worth paying more efforts to achieve better performance.

eess.AS↗

arXiv2020

Detecting Early-warning signals in Time Series of Visits to Points of Interests to Examine Population Response to COVID -19 Pandemic

The objective of this paper is to examine population response to COVID-19 and associated policy interventions through detecting early-warning signals in time series of visits to points of interest (POIs). Complex systems, such as cities, demonstrate early-warning signals when they approach phase transitions responding to external perturbation, including crises, policy changes, and human behavior changes. In urban systems, population visits to POIs represent a state in the complex systems that are cities. These states may undergo phase transitions due to population response to pandemic risks and intervention policies. In this study, we conducted early-warning signal detection on population visits to POIs to examine population response to pandemic risks. We examined two early-warning signals, the increase of autocorrelation at-lag-1 and standard deviation, in time series of population visits to POIs in 17 metropolitan cities in the United States of America. The results show that: (1) early-warning signals for population response to COVID-19 were detected between February 14 and March 11, 2020 in 17 cities; (2) detected population response had started prior to shelter-in-place orders in 17 cities; (3) early-warning signals detected from the essential POIs visits appeared earlier than those from non-essential POIs; and 4) longer time lags between detected population response and shelter-in-place orders led to a less decrease in POI visits. The results show the importance of detecting early-warning signals during crises in cities as complex systems. Early-warning signals could provide important insights regarding the timing and extent of population response to crises to inform policy makers.

physics.soc-ph↗

arXiv2020

ASR-Free Pronunciation Assessment

Most of the pronunciation assessment methods are based on local features derived from automatic speech recognition (ASR), e.g., the Goodness of Pronunciation (GOP) score. In this paper, we investigate an ASR-free scoring approach that is derived from the marginal distribution of raw speech signals. The hypothesis is that even if we have no knowledge of the language (so cannot recognize the phones/words), we can still tell how good a pronunciation is, by comparatively listening to some speech data from the target language. Our analysis shows that this new scoring approach provides an interesting correction for the phone-competition problem of GOP. Experimental results on the ERJ dataset demonstrated that combining the ASR-free score and GOP can achieve better performance than the GOP baseline.

eess.AS↗

arXiv2020

Distributed Optimal Generation and Load-Side Control for Frequency Regulation in Power Systems

In order to deal with issues caused by the increasing penetration of renewable resources in power systems, this paper proposes a novel distributed frequency control algorithm for each generating unit and controllable load in a transmission network to replace the conventional automatic generation control (AGC). The targets of the proposed control algorithm are twofold. First, it is to restore the nominal frequency and scheduled net inter-area power exchanges after an active power mismatch between generation and demand. Second, it is to optimally coordinate the active powers of all controllable units in a distributed manner. The designed controller only relies on local information, computation, and peer-to-peer communication between cyber-connected buses, and it is also robust against uncertain system parameters. Asymptotic stability of the closed-loop system under the designed algorithm is analysed by using a nonlinear structure-preserving model including the first-order turbine-governor dynamics. Finally, case studies validate the effectiveness of the proposed method.

eess.SY↗

arXiv2019

AP19-OLR Challenge: Three Tasks and Their Baselines

This paper introduces the fourth oriental language recognition (OLR) challenge AP19-OLR, including the data profile, the tasks and the evaluation principles. The OLR challenge has been held successfully for three consecutive years, along with APSIPA Annual Summit and Conference (APSIPA ASC). The challenge this year still focuses on practical and challenging tasks, precisely (1) short-utterance LID, (2) cross-channel LID and (3) zero-resource LID. The event this year includes more languages and more real-life data provided by SpeechOcean and the NSFC M2ASR project. All the data is free for participants. Recipes for x-vector system and back-end evaluation are also conducted as baselines for the three tasks. The participants can refer to these online-published recipes to deploy LID systems for convenience. We report the baseline results on the three tasks and demonstrate that the three tasks are worth some efforts to achieve better performance.

eess.AS↗

arXiv2019

Distributed Inter-Area Oscillation Damping Control for Power Systems by Using Wind Generators and Load Aggregators

This paper investigates the potential of wind turbine generators (WTGs) and load aggregators (LAs) to provide supplementary damping control services for low frequency inter-area oscillations (LFOs) through the additional distributed damping control units (DCUs) proposed in their controllers. In order to provide a scalable methodology for the increasing number of WTGs and LAs, a novel distributed control framework is proposed to coordinate damping controllers. Firstly, a distributed algorithm is designed to reconstruct the system Jacobian matrix for each damping bus (buses with damping controllers). Thus, the critical LFO can be identified locally at each damping bus by applying eigen-analysis to the obtained system Jacobian matrix. Then, if the damping ratio of the critical LFO is less than a preset threshold, the control parameters of DCUs will be tuned in a distributed and coordinated manner to improve the damping ratio and minimize the total control cost at the same time. The proposed control framework is tested in a modified IEEE 39-bus test system. The simulation results with and without the proposed control framework are compared to demonstrate the effectiveness of the proposed framework.

math.OC↗

arXiv2019

Gaussian-Constrained training for speaker verification

Neural models, in particular the d-vector and x-vector architectures, have produced state-of-the-art performance on many speaker verification tasks. However, two potential problems of these neural models deserve more investigation. Firstly, both models suffer from `information leak', which means that some parameters participating in model training will be discarded during inference, i.e, the layers that are used as the classifier. Secondly, these models do not regulate the distribution of the derived speaker vectors. This `unconstrained distribution' may degrade the performance of the subsequent scoring component, e.g., PLDA. This paper proposes a Gaussian-constrained training approach that (1) discards the parametric classifier, and (2) enforces the distribution of the derived speaker vectors to be Gaussian. Our experiments on the VoxCeleb and SITW databases demonstrated that this new training approach produced more representative and regular speaker embeddings, leading to consistent performance improvement.

eess.AS↗

arXiv2018

Phonetic-attention scoring for deep speaker features in speaker verification

Recent studies have shown that frame-level deep speaker features can be derived from a deep neural network with the training target set to discriminate speakers by a short speech segment. By pooling the frame-level features, utterance-level representations, called d-vectors, can be derived and used in the automatic speaker verification (ASV) task. This simple average pooling, however, is inherently sensitive to the phonetic content of the utterance. An interesting idea borrowed from machine translation is the attention-based mechanism, where the contribution of an input word to the translation at a particular time is weighted by an attention score. This score reflects the relevance of the input word and the present translation. We can use the same idea to align utterances with different phonetic contents. This paper proposes a phonetic-attention scoring approach for d-vector systems. By this approach, an attention score is computed for each frame pair. This score reflects the similarity of the two frames in phonetic content, and is used to weigh the contribution of this frame pair in the utterance-based scoring. This new scoring approach emphasizes the frame pairs with similar phonetic contents, which essentially provides a soft alignment for utterances with any phonetic contents. Experimental results show that compared with the naive average pooling, this phonetic-attention scoring approach can deliver consistent performance improvement in ASV tasks of both text-dependent and text-independent.

eess.AS↗

arXiv2018

AP18-OLR Challenge: Three Tasks and Their Baselines

The third oriental language recognition (OLR) challenge AP18-OLR is introduced in this paper, including the data profile, the tasks and the evaluation principles. Following the events in the last two years, namely AP16-OLR and AP17-OLR, the challenge this year focuses on more challenging tasks, including (1) short-duration utterances, (2) confusing languages, and (3) open-set recognition. The same as the previous events, the data of AP18-OLR is also provided by SpeechOcean and the NSFC M2ASR project. Baselines based on both the i-vector model and neural networks are constructed for the participants' reference. We report the baseline results on the three tasks and demonstrate that the three tasks are truly challenging. All the data is free for participants, and the Kaldi recipes for the baselines have been published online.

cs.CL↗

arXiv2018

Full-info Training for Deep Speaker Feature Learning

In recent studies, it has shown that speaker patterns can be learned from very short speech segments (e.g., 0.3 seconds) by a carefully designed convolutional & time-delay deep neural network (CT-DNN) model. By enforcing the model to discriminate the speakers in the training data, frame-level speaker features can be derived from the last hidden layer. In spite of its good performance, a potential problem of the present model is that it involves a parametric classifier, i.e., the last affine layer, which may consume some discriminative knowledge, thus leading to `information leak' for the feature learning. This paper presents a full-info training approach that discards the parametric classifier and enforces all the discriminative knowledge learned by the feature net. Our experiments on the Fisher database demonstrate that this new training scheme can produce more coherent features, leading to consistent and notable performance improvement on the speaker verification task.

cs.SD↗

arXiv2018

Deep factorization for speech signal

Various informative factors mixed in speech signals, leading to great difficulty when decoding any of the factors. An intuitive idea is to factorize each speech frame into individual informative factors, though it turns out to be highly difficult. Recently, we found that speaker traits, which were assumed to be long-term distributional properties, are actually short-time patterns, and can be learned by a carefully designed deep neural network (DNN). This discovery motivated a cascade deep factorization (CDF) framework that will be presented in this paper. The proposed framework infers speech factors in a sequential way, where factors previously inferred are used as conditional variables when inferring other factors. We will show that this approach can effectively factorize speech signals, and using these factors, the original speech spectrum can be recovered with a high accuracy. This factorization and reconstruction approach provides potential values for many speech processing tasks, e.g., speaker recognition and emotion recognition, as will be demonstrated in the paper.

eess.AS↗