arXiv subjects

Yonghong Yan

Publications and source records attributed to Yonghong Yan.

At least 37 records · Page 2Linked to original sources

arXiv2022

Open Source MagicData-RAMC: A Rich Annotated Mandarin Conversational(RAMC) Speech Dataset

This paper introduces a high-quality rich annotated Mandarin conversational (RAMC) speech dataset called MagicData-RAMC. The MagicData-RAMC corpus contains 180 hours of conversational speech data recorded from native speakers of Mandarin Chinese over mobile phones with a sampling rate of 16 kHz. The dialogs in MagicData-RAMC are classified into 15 diversified domains and tagged with topic labels, ranging from science and technology to ordinary life. Accurate transcription and precise speaker voice activity timestamps are manually labeled for each sample. Speakers' detailed information is also provided. As a Mandarin speech dataset designed for dialog scenarios with high quality and rich annotations, MagicData-RAMC enriches the data diversity in the Mandarin speech community and allows extensive research on a series of speech-related tasks, including automatic speech recognition, speaker diarization, topic detection, keyword search, text-to-speech, etc. We also conduct several relevant tasks and provide experimental results to help evaluate the dataset.

cs.CL↗

arXiv2022

A Novel Blind Source Separation Framework Towards Maximum Signal-To-Interference Ratio

This letter proposes a new blind source separation (BSS) framework termed minimum variance independent component analysis (MVICA), which can potentially achieve the maximum output signal-to-interference ratio (SIR) while also allowing more flexibility in real implementations. The statistical independence assumption has been the foundation of the most dominant BSS techniques in recent decades. However, this assumption does not always hold true and the accurate probabilistic modeling of source is inherently difficult. To overcome these limitations and improve the separation performance, the MVICA framework is rigorously derived by optimizing the design of these independence-based BSS algorithms with the maximum SIR criterion. A deep neural network-supported implementation of MVICA is subsequently described. Experimental results under various conditions show the superiority of MVICA over the state-of-the-art BSS algorithms, in terms of not only SIR but also signal-to-distortion ratio and automatic speech recognition rate.

cs.SD↗

arXiv2021

Decomposing Complex Questions Makes Multi-Hop QA Easier and More Interpretable

Multi-hop QA requires the machine to answer complex questions through finding multiple clues and reasoning, and provide explanatory evidence to demonstrate the machine reasoning process. We propose Relation Extractor-Reader and Comparator (RERC), a three-stage framework based on complex question decomposition, which is the first work that the RERC model has been proposed and applied in solving the multi-hop QA challenges. The Relation Extractor decomposes the complex question, and then the Reader answers the sub-questions in turn, and finally the Comparator performs numerical comparison and summarizes all to get the final answer, where the entire process itself constitutes a complete reasoning evidence path. In the 2WikiMultiHopQA dataset, our RERC model has achieved the most advanced performance, with a winning joint F1 score of 53.58 on the leaderboard. All indicators of our RERC are close to human performance, with only 1.95 behind the human level in F1 score of support fact. At the same time, the evidence path provided by our RERC framework has excellent readability and faithfulness.

cs.CL↗

arXiv2021

Reminding the Incremental Language Model via Data-Free Self-Distillation

Incremental language learning with pseudo-data can alleviate catastrophic forgetting in neural networks. However, to obtain better performance, former methods have higher demands for pseudo-data of the previous tasks. The performance dramatically decreases when fewer pseudo-data are employed. In addition, the distribution of pseudo-data gradually deviates from the real data with the sequential learning of different tasks. The deviation will be greater with more tasks learned, which results in more serious catastrophic forgetting. To address these issues, we propose reminding incremental language model via data-free self-distillation (DFSD), which includes self-distillation based on the Earth Mover's Distance and hidden data augmentation. By estimating the knowledge distribution in all layers of GPT-2 and transforming it from teacher model to student model, the Self-distillation based on the Earth Mover's Distance can significantly reduce the demand for pseudo-data. Hidden data augmentation can greatly alleviate the catastrophic forgetting caused by deviations via modeling the generation of pseudo-data as a hidden data augmentation process, where each sample is a mixture of all trained task data. The experimental results demonstrate that our DFSD can exceed the previous state-of-the-art methods even if the maximum decrease in pseudo-data is 90%.

cs.CL↗

arXiv2021

The HCCL Speaker Verification System for Far-Field Speaker Verification Challenge

This paper describes the systems submitted by team HCCL to the Far-Field Speaker Verification Challenge. Our previous work in the AIshell Speaker Verification Challenge 2019 shows that the powerful modeling abilities of Neural Network architectures can provide exceptional performance for this kind of task. Therefore, in this challenge, we focus on constructing deep Neural Network architectures based on TDNN, Resnet and Res2net blocks. Most of the developed systems consist of Neural Network embeddings are applied with PLDA backend. Firstly, the speed perturbation method is applied to augment data and significant performance improvements are achieved. Then, we explore the use of AMsoftmax loss function and propose to join a CE-loss branch when we train model using AMsoftmax loss. In addition, the impact of score normalization on performance is also investigated. The final system, a fusion of four systems, achieves minDCF 0.5342, EER 5.05\% on task1 eval set, and achieves minDCF 0.5193, EER 5.47\% on task3 eval set.

cs.SD↗

arXiv2021

Improved Conformer-based End-to-End Speech Recognition Using Neural Architecture Search

Recently neural architecture search(NAS) has been successfully used in image classification, natural language processing, and automatic speech recognition(ASR) tasks for finding the state-of-the-art(SOTA) architectures than those human-designed architectures. NAS can derive a SOTA and data-specific architecture over validation data from a pre-defined search space with a search algorithm. Inspired by the success of NAS in ASR tasks, we propose a NAS-based ASR framework containing one search space and one differentiable search algorithm called Differentiable Architecture Search(DARTS). Our search space follows the convolution-augmented transformer(Conformer) backbone, which is a more expressive ASR architecture than those used in existing NAS-based ASR frameworks. To improve the performance of our method, a regulation method called Dynamic Search Schedule(DSS) is employed. On a widely used Mandarin benchmark AISHELL-1, our best-searched architecture outperforms the baseline Conform model significantly with about 11% CER relative improvement, and our method is proved to be pretty efficient by the search cost comparisons.

eess.AS↗

arXiv2020

Multi-Accent Adaptation based on Gate Mechanism

When only a limited amount of accented speech data is available, to promote multi-accent speech recognition performance, the conventional approach is accent-specific adaptation, which adapts the baseline model to multiple target accents independently. To simplify the adaptation procedure, we explore adapting the baseline model to multiple target accents simultaneously with multi-accent mixed data. Thus, we propose using accent-specific top layer with gate mechanism (AST-G) to realize multi-accent adaptation. Compared with the baseline model and accent-specific adaptation, AST-G achieves 9.8% and 1.9% average relative WER reduction respectively. However, in real-world applications, we can't obtain the accent category label for inference in advance. Therefore, we apply using an accent classifier to predict the accent label. To jointly train the acoustic model and the accent classifier, we propose the multi-task learning with gate mechanism (MTL-G). As the accent label prediction could be inaccurate, it performs worse than the accent-specific adaptation. Yet, in comparison with the baseline model, MTL-G achieves 5.1% average relative WER reduction.

eess.AS↗

arXiv2020

A Model Compression Method with Matrix Product Operators for Speech Enhancement

The deep neural network (DNN) based speech enhancement approaches have achieved promising performance. However, the number of parameters involved in these methods is usually enormous for the real applications of speech enhancement on the device with the limited resources. This seriously restricts the applications. To deal with this issue, model compression techniques are being widely studied. In this paper, we propose a model compression method based on matrix product operators (MPO) to substantially reduce the number of parameters in DNN models for speech enhancement. In this method, the weight matrices in the linear transformations of neural network model are replaced by the MPO decomposition format before training. In experiment, this process is applied to the causal neural network models, such as the feedforward multilayer perceptron (MLP) and long short-term memory (LSTM) models. Both MLP and LSTM models with/without compression are then utilized to estimate the ideal ratio mask for monaural speech enhancement. The experimental results show that our proposed MPO-based method outperforms the widely-used pruning method for speech enhancement under various compression rates, and further improvement can be achieved with respect to low compression rates. Our proposal provides an effective model compression method for speech enhancement, especially in cloud-free application.

cs.SD↗

arXiv2020

Transformer-based Online CTC/attention End-to-End Speech Recognition Architecture

Recently, Transformer has gained success in automatic speech recognition (ASR) field. However, it is challenging to deploy a Transformer-based end-to-end (E2E) model for online speech recognition. In this paper, we propose the Transformer-based online CTC/attention E2E ASR architecture, which contains the chunk self-attention encoder (chunk-SAE) and the monotonic truncated attention (MTA) based self-attention decoder (SAD). Firstly, the chunk-SAE splits the speech into isolated chunks. To reduce the computational cost and improve the performance, we propose the state reuse chunk-SAE. Sencondly, the MTA based SAD truncates the speech features monotonically and performs attention on the truncated features. To support the online recognition, we integrate the state reuse chunk-SAE and the MTA based SAD into online CTC/attention architecture. We evaluate the proposed online models on the HKUST Mandarin ASR benchmark and achieve a 23.66% character error rate (CER) with a 320 ms latency. Our online model yields as little as 0.19% absolute CER degradation compared with the offline baseline, and achieves significant improvement over our prior work on Long Short-Term Memory (LSTM) based online E2E models.

eess.AS↗

arXiv2019

Multi-Talker MVDR Beamforming Based on Extended Complex Gaussian Mixture Model

In this letter, we present a novel multi-talker minimum variance distortionless response (MVDR) beamforming as the front-end of an automatic speech recognition (ASR) system in a dinner party scenario. The CHiME-5 dataset is selected to evaluate our proposal for overlapping multi-talker scenario with severe noise. A detailed study on beamforming is conducted based on the proposed extended complex Gaussian mixture model (CGMM) integrated with various speech separation and speech enhancement masks. Three main changes are made to adopt the original CGMM-based MVDR for the multi-talker scenario. First, the number of Gaussian distributions is extended to 3 with an additional inference speaker model. Second, the mixture coefficients are introduced as a supervisor to generate more elaborate masks and avoid the permutation problems. Moreover, we reorganize the MVDR and mask-based speech separation to achieve both noise reduction and target speaker extraction. With the official baseline ASR back-end, our front-end algorithm gained an absolute WER reduction of 13.87% compared with the baseline front-end.

eess.AS↗

arXiv2019

Integrating the Data Augmentation Scheme with Various Classifiers for Acoustic Scene Modeling

This technical report describes the IOA team's submission for TASK1A of DCASE2019 challenge. Our acoustic scene classification (ASC) system adopts a data augmentation scheme employing generative adversary networks. Two major classifiers, 1D deep convolutional neural network integrated with scalogram features and 2D fully convolutional neural network integrated with Mel filter bank features, are deployed in the scheme. Other approaches, such as adversary city adaptation, temporal module based on discrete cosine transform and hybrid architectures, have been developed for further fusion. The results of our experiments indicates that the final fusion systems A-D could achieve an accuracy higher than 85% on the officially provided fold 1 evaluation dataset.

eess.AS↗

arXiv2018

A Cross-Layer Solution in Scientific Workflow System for Tackling Data Movement Challenge

Scientific applications in HPC environment are more com-plex and more data-intensive nowadays. Scientists usually rely on workflow system to manage the complexity: simply define multiple processing steps into a single script and let the work-flow systems compile it and schedule all tasks accordingly. Numerous workflow systems have been proposed and widely used, like Galaxy, Pegasus, Taverna, Kepler, Swift, AWE, etc., to name a few examples. Traditionally, scientific workflow systems work with parallel file systems, like Lustre, PVFS, Ceph, or other forms of remote shared storage systems. As such, the data (including the intermediate data generated during workflow execution) need to be transferred back and forth between compute nodes and storage systems, which introduces a significant performance bottleneck on I/O operations. Along with the enlarging perfor-mance gap between CPU and storage devices, this bottleneck is expected to be worse. Recently, we have introduced a new concept of Compute-on-Data-Path to allow tasks and data binding to be more efficient to reduce the data movement cost. To workflow systems, the key is to exploit the data locality in HPC storage hierarchy: if the datasets are stored in compute nodes, near the workflow tasks, then the task can directly access them with better performance with less network usage. Several recent studies have been done regarding building such a shared storage system, utilizing compute node resources, to serve HPC workflows with locality, such as Hercules [1] and WOSS [2] etc. In this research, we further argue that providing a compute-node side storage system is not sufficient to fully exploit data locality. A cross-layer solution combining storage system, compiler, and runtime is necessary. We take Swift/T [3], a workflow system for data-intensive applications, as a prototype platform to demonstrate such a cross-layer solution

cs.DC↗

arXiv2017

Stream Attention for far-field multi-microphone ASR

A stream attention framework has been applied to the posterior probabilities of the deep neural network (DNN) to improve the far-field automatic speech recognition (ASR) performance in the multi-microphone configuration. The stream attention scheme has been realized through an attention vector, which is derived by predicting the ASR performance from the phoneme posterior distribution of individual microphone stream, focusing the recognizer's attention to more reliable microphones. Investigation on the various ASR performance measures has been carried out using the real recorded dataset. Experiments results show that the proposed framework has yielded substantial improvements in word error rate (WER).

cs.SD↗

arXiv2017

Rank-1 Constrained Multichannel Wiener Filter for Speech Recognition in Noisy Environments

Multichannel linear filters, such as the Multichannel Wiener Filter (MWF) and the Generalized Eigenvalue (GEV) beamformer are popular signal processing techniques which can improve speech recognition performance. In this paper, we present an experimental study on these linear filters in a specific speech recognition task, namely the CHiME-4 challenge, which features real recordings in multiple noisy environments. Specifically, the rank-1 MWF is employed for noise reduction and a new constant residual noise power constraint is derived which enhances the recognition performance. To fulfill the underlying rank-1 assumption, the speech covariance matrix is reconstructed based on eigenvectors or generalized eigenvectors. Then the rank-1 constrained MWF is evaluated with alternative multichannel linear filters under the same framework, which involves a Bidirectional Long Short-Term Memory (BLSTM) network for mask estimation. The proposed filter outperforms alternative ones, leading to a 40% relative Word Error Rate (WER) reduction compared with the baseline Weighted Delay and Sum (WDAS) beamformer on the real test set, and a 15% relative WER reduction compared with the GEV-BAN method. The results also suggest that the speech recognition accuracy correlates more with the Mel-frequency cepstral coefficients (MFCC) feature variance than with the noise reduction or the speech distortion level.

cs.SD↗

arXiv2017

Relative Transfer Function Inverse Regression from Low Dimensional Manifold

In room acoustic environments, the Relative Transfer Functions (RTFs) are controlled by few underlying modes of variability. Accordingly, they are confined to a low-dimensional manifold. In this letter, we investigate a RTF inverse regression problem, the task of which is to generate the high-dimensional responses from their low-dimensional representations. The problem is addressed from a pure data-driven perspective and a supervised Deep Neural Network (DNN) model is applied to learn a mapping from the source-receiver poses (positions and orientations) to the frequency domain RTF vectors. The experiments show promising results: the model achieves lower prediction error of the RTF than the free field assumption. However, it fails to compete with the linear interpolation technique in small sampling distances.

cs.SD↗

arXiv2016

Optimizing human-interpretable dialog management policy using Genetic Algorithm

Automatic optimization of spoken dialog management policies that are robust to environmental noise has long been the goal for both academia and industry. Approaches based on reinforcement learning have been proved to be effective. However, the numerical representation of dialog policy is human-incomprehensible and difficult for dialog system designers to verify or modify, which limits its practical application. In this paper we propose a novel framework for optimizing dialog policies specified in domain language using genetic algorithm. The human-interpretable representation of policy makes the method suitable for practical employment. We present learning algorithms using user simulation and real human-machine dialogs respectively.Empirical experimental results are given to show the effectiveness of the proposed approach.

cs.HC↗

arXiv2015

Noise Robust IOA/CAS Speech Separation and Recognition System For The Third 'CHIME' Challenge

This paper presents the contribution to the third 'CHiME' speech separation and recognition challenge including both front-end signal processing and back-end speech recognition. In the front-end, Multi-channel Wiener filter (MWF) is designed to achieve background noise reduction. Different from traditional MWF, optimized parameter for the tradeoff between noise reduction and target signal distortion is built according to the desired noise reduction level. In the back-end, several techniques are taken advantage to improve the noisy Automatic Speech Recognition (ASR) performance including Deep Neural Network (DNN), Convolutional Neural Network (CNN) and Long short-term memory (LSTM) using medium vocabulary, Lattice rescoring with a big vocabulary language model finite state transducer, and ROVER scheme. Experimental results show the proposed system combining front-end and back-end is effective to improve the ASR performance.

cs.SD↗

arXiv2014

Energy spread and current-current correlation in quantum systems

We consider energy (heat) transport in quantum systems, and establish a relationship between energy spread and energy current-current correlation function. The energy current-current correlation is related to thermal conductivity by the Green-Kubo formula, and thus this relationship allows us to study conductivity directly from the energy spread process. As an example, we investigate a spinless fermion model; the numerical results confirm the relationship.

cond-mat.stat-mech↗