检索结果-内蒙古大学图书馆

14th IEEE International Conference on Signal processing, Communications and Computing, ICSPCC 2024

作者： Yang, Xue Bao, Changchun Zhou, Jing Zhang, Xu Duan, Haiwei Zhao, Yunhao Li, Wenwen Institute of Speech and Audio Information Processing Beijing University of Technology Faculty of Information Technology China

ISBN: (纸本)9798350366556

Multichannel speech processing has been widely studied since the spatial information contained in the multichannel signals can be exploited. To facilitate the efficient transmission and preserve the spatial information, the multichannel speech coding technique is needed. Recently, a multichannel speech coding method based on the Opus codec and spatial parameters was proposed. In the encoding stage, the speech signal of the reference channel is encoded with the Opus codec. The multichannel speech signals are decomposed through the Gammatone filter bank and the spatial parameters are extracted and quantized for each sub-band signals. In the decoding stage, the encoded signal of the reference channel is decoded with the Opus codec. Subsequently, this decoded signal is combined with the quantized spatial parameters to recover the speech signals of the remaining channels. In this paper, an improved implementation of this coding method is detailed. Specifically, a framing pattern more suitable for multichannel speech coding with the Gammatone filter bank is proposed. Besides, a newly designed window is then employed on each sub-band signal for the precise extraction of spatial parameters in the frequency domain. Additionally, the extracted spatial parameters are quantized non-uniformly. The experimental results show the effectiveness of the proposed implementation. This implementation can achieve high speech quality with a reduced bitrate. Furthermore, the spatial information contained in the multichannel speech signals can be better preserved. © 2024 IEEE.

关键词： Frequency domain analysis

来源：评论

学校读者我要写书评

暂无评论

Time-Frequency Generative Adversarial Network with Transformer for Packet Loss Concealment 17

Time-Frequency Generative Adversarial Network with Transform...

引用

17th IEEE International Conference on Signal processing, ICSP 2024

作者： Zhao, Yunhao Bao, Changchun Yang, Xue Zhang, Xu Institute of Speech and Audio Information Processing School of Information Science and Technology Beijing China

ISBN: (纸本)9798350387384

With the advancement of internet technology, internet communication has become an indispensable part in our daily life. However, in actual conversations, due to the latency and jitter of the network, the data packet may be lost and have a negative impact on the quality of communication. This paper introduces a generative adversarial network with attention mechanism, which effectively restores the missing speech by the already received historical information. The generator composed of dilated convolution and Transformer module is employed to extract both local and global features from the historical information and generate the missed speech. The discriminator is composed of two multi-resolution discriminative networks for further enhancing the accuracy of the recovered speech. Experimental validation demonstrates the effectiveness of this approach in restoring lost speech, and the communication quality in the occurrence of packet loss is enhanced. © 2024 IEEE.

关键词： Packet loss

来源：评论

学校读者我要写书评

暂无评论

Weighting Function Modification Used for Phase Transform-Based Time Delay Estimation

引用

China Communications 2022年第11期19卷 241-256页

作者： Xue Yang Changchun Bao Zihao Cui Speech and Audio Signal Processing Laboratory Faculty of Information TechnologyBeijing University of TechnologyBeijing 100124China

Generalized cross-correlation is considered as the most straightforward time delay estimation *** on various weighting function,different methods were derived and a straightforward method,named phase transform(PHAT)has been widely *** is well-known for its robustness to reverberation and its sensitivity to noise,which is partly due to the fact that PHAT distributes same weights to the frequencies dominated by signal or *** alleviate this problem,two weighting functions are proposed in this *** taking a posteriori signal-to-noise ratio(SNR)into account to classify reliable and unreliable frequencies,different weights could be *** first proposed weighting function borrows the idea of binary mask and distributes same weights to frequencies in same set,whereas,the second one assigns weights based on coherence *** showed the robustness of proposed methods to reverberation and noise for improving the performance of time delay estimation through various criteria.

关键词： time delay estimation generalized crosscorrelation PHAT a posteriori SNR coherence function

来源：评论

学校读者我要写书评

暂无评论

A Packet Loss Concealment Method Based on the Demucs Network Structure 1

引用

18th National Conference on Man-Machine Speech Communication, NCMMSC 2023

作者： Li, Wenwen Bao, Changchun Institute of Speech and Audio Information Processing Faculty of Information Technology Beijing University of Technology Beijing100124 China

ISBN: (数字)9789819706013

ISBN: (纸本)9789819706006

Under the constrained real-time condition of network, packet loss often occurs in voice communications like voice over internet protocol (VoIP). Packet loss concealment (PLC) techniques often use the previous packets to recover the lost packet for improving the quality of speech communication. In this paper, a novel deep PLC approach is proposed, which uses a called Demucs network structure, i.e., a deep U-Net architecture with a long-short time memory (LSTM) network, to predict the lost packet in the time domain. Firstly, by combing the convolutions with gated linear unit (GLU), the encoder of network can systematically extract the high-level feature of each speech frame. Secondly, the LSTM layers are used to learn the long-term dependencies of speech frames. Finally, the U-Net architecture of the network is used to improve the gradient of information flow by using skip connections, which enhances the decoder’s ability of reconstructing the lost speech frames. Additionally, the proposed architecture is optimized by utilizing multiple loss functions in the time and frequency domains. The experimental results show that the proposed method has better performance in perceptual evaluation of speech quality (PESQ) and short-term objective intelligibility (STOI). © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Packet loss

来源：评论

学校读者我要写书评

暂无评论

A Simplified Post-Training Quantization Method for Multi-Channel Speech Separation Model 17

A Simplified Post-Training Quantization Method for Multi-Cha...

引用

17th IEEE International Conference on Signal processing, ICSP 2024

作者： Zhang, Xu Bao, Changchun Yang, Xue Beijing University of Technology Institute of Speech and Audio Information Processing School of Information Science and Technology Beijing100124 China

ISBN: (纸本)9798350387384

In recent years, the performance of speech separation has been greatly improved by using neural networks. However, the increasing complexity of neural network design usually requires a large amount of computing resources and storage space, which limits its application in resource-constrained environments. To address this issue, the model compression techniques have emerged. By quantizing the weights and activation values of the model, the computational cost can be significantly reduced while maintaining separation performance. This paper introduces a simplified post-training quantization (PTQ) method to compress existing speech separation models by using 8-bit precision for both weight and activation instead of 32-bit precision. The experiments on the spatialized version of the WSJ0-2mix corpus show that the proposed method keeps comparable performance with the baseline speech separation systems in terms of SDR scores, while moderately reduces the complexity. © 2024 IEEE.

关键词： Source separation

来源：评论

学校读者我要写书评

暂无评论

Binaural detection thresholds are differently affected by interaural level differences in S0 and Sπ conditions 10

Binaural detection thresholds are differently affected by in...

引用

10th Convention of the European Acoustics Association, EAA 2023

作者： Bischof, Norbert F. Seeber, Bernhard U. Audio Information Processing School of Computation Information and Technology Technical University of Munich Germany

ISBN: (纸本)9788888942674

The dependency of binaural detection thresholds on interaural level differences was investigated. Detection thresholds were measured for four discrete tones with either 100, 500, 1000 or 2500 Hz presented in diotic, uniform exciting noise (N0) via headphones using an adaptive forced choice method. The sine tones were presented either in-phase (S0) or anti-phasic (Sπ), and the interaural level difference (ILD) of the tone was varied within 15 dB. Preliminary detection thresholds show that unmasking depends on the applied ILD. In the N0S0 condition, thresholds of low-frequency tones decrease by up to 7 dB for an ILD of 15 dB, which almost corresponds to the monaural threshold. This decrease with increasing ILD cannot be observed at 2500 Hz. In the N0Sπ condition, however, the increase in thresholds with increasing ILD is approximately equal for all frequencies. The resulting decrease in unmasking seems to follow the level decrease of the two summed sine tones with increasing ILD. In particular, the results of the N0Sπ condition suggest that the amount of binaural unmasking is more affected by the ILD than in the N0S0 condition. © 2023 Norbert F Bischof and Bernhard U. Seeber. This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 Unported License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

关键词： Audition

来源：评论

学校读者我要写书评

暂无评论

Multi-Source Localization Method Based on the Log-Mel Spectrum Augmented Noise Subspace

Multi-Source Localization Method Based on the Log-Mel Spectr...

引用

2023 IEEE International Conference on Signal processing, Communications and Computing, ICSPCC 2023

作者： Duan, Haiwei Bao, Changchun Zhou, Jing Beijing University of Technology Speech and Audio Signal Processing Laboratory Faculty of Information Technology Beijing China

ISBN: (纸本)9798350316728

The deep learning (DL) based direction-of-arrival (DOA) estimation is one of the research hotspots, and many methods have been proposed recently. However, most of those methods will face serious performance degradation, since the adverse impacts caused by the sources overlapping, noise and reverberation. One of the primary impacts is that the performance degradation is susceptible to some pre-extracted features that often result in spectral aliasing and peak confusion in a complex scenario. In this paper, a new feature stacked with the log-Mel spectrum and the noise subspace of the covariance matrix of the relative sound pressure is proposed and further used for the DL-based DOA estimation, which is referred to log-Mel spectrum augmented noise subspace (LMNS). The LMNS is more robust compared with the conventional features since it can represent both spectral and spatial information effectively. Meanwhile, the LMNS is used as the input feature and fed to a Conformer based residual network to map the spatial pseudo-spectrum, thereby the DOAs of the sound sources can be obtained. The experimental results show that the proposed method has better performance on the DOA estimation, which verifies that the proposed feature LMNS is more robust and effective in the scenarios with multi-source, noise and reverberation. © 2023 IEEE.

关键词： Reverberation

来源：评论

学校读者我要写书评

暂无评论

Predictive Packet Loss Concealment Method Based on Conformer and Temporal Convolution Module

Predictive Packet Loss Concealment Method Based on Conformer...

引用

2023 IEEE International Conference on Signal processing, Communications and Computing, ICSPCC 2023

作者： Zhao, Yunhao Bao, Changchun Yang, Xue Beijing University of Technology Speech and Audio Signal Processing Laboratory Faculty of Information Technology Beijing100124 China

ISBN: (纸本)9798350316728

To enhance the perceptual quality of speech signal, the Packet Loss Concealment (PLC) technique focuses on recovering the lost speech caused by network latency and jitter. In practical applications, the PLC methods typically employ a predictive process that relies on previously received speech signal to recover the lost speech without introducing additional delay. In this paper, we propose a predictive PLC network, which employs the Conformer and temporal convolution module to fully exploit the contextual dependencies and to better predict the lost speech. In addition, our proposed network can be directly employed as the generator and combined with appropriate discriminative networks, forming a Generative Adversarial Network (GAN) paradigm that can enhance the perceptual quality of the recovered speech signal. Experimental results demonstrate that without any discriminative network, the proposed method demonstrates impressive results in PLC. Under the GAN paradigm, further improvement can be observed and our proposed method outperforms several baseline methods at different packet loss rates. © 2023 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

A Beam-TFDPRNN Based Speech Separation Method in Reverberant Environments

A Beam-TFDPRNN Based Speech Separation Method in Reverberant...

引用

2023 IEEE International Conference on Signal processing, Communications and Computing, ICSPCC 2023

作者： Zhang, Xu Bao, Changchun Zhou, Jing Yang, Xue Beijing University of Technology Speech and Audio Signal Processing Laboratory Faculty of Information Technology Beijing100124 China

ISBN: (纸本)9798350316728

Recently, the beamforming methods based on the time domain audio separation network (Beam-TasNet) have shown satisfactory performance. For example, the performance of minimum variance distortionless response (MVDR) beamformer can be effectively improved by using the time domain audio separation network (TasNet). However, the reverberation will draw a significant performance degradation to the Beam-TasNet since the multiple reflection sounds damage the time domain features extracted by the TasNet. Fortunately, the recent studies show that the frequency domain features have better anti-interference capability in the reverberation environment. Therefore, this paper proposed a MVDR beamforming method based on the time-frequency domain Dual-Path Recurrent Neural Network (TFDPRNN) for the task of speech separation, and we call it Beam-TFDPRNN. In this method, the TFDPRNN uses a path scanning mechanism to capture the time-frequency features more comprehensively by repeatedly scanning input speech signal in both the time and frequency dimensions. The scanned time-frequency features could describe the characteristics of the speech sources in reverberation environment more robustly. As a result, a better pre-separation result is obtained by the TFDPRNN. Furthermore, by using the pre-separation signals to calculate the spatial covariance matrices, a more robust MVDR beamformer is obtained so that the speech separation in the reverberation environment is achieved efficiently. The experiment results based on the WSJ0-2mix corpus show that the proposed method achieves superior performance compared to the reference methods. © 2023 IEEE.

关键词： Reverberation

来源：评论

学校读者我要写书评

暂无评论

A Time-domain Packet Loss Concealment Method by Designing Transformer-based Convolutional Recurrent Network

A Time-domain Packet Loss Concealment Method by Designing Tr...

引用

2023 IEEE International Conference on Signal processing, Communications and Computing, ICSPCC 2023

作者： Li, Wenwen Bao, Changchun Zhou, Jing Yang, Xue Beijing University of Technology Speech and Audio Signal Processing Laboratory Faculty of Information Technology Beijing 100124 China

ISBN: (纸本)9798350316728

Due to the network constrain, the packet loss is inevitable in the real-Time speech communication. The packet loss often leads to the short interruption of the voice communication, which seriously affects the quality of speech communication. In this paper, a transformer-based convolutional recurrent network (CRN-Trans) is designed for packet loss concealment (PLC) in time domain. In the CRN-Tans, the convolutional layers are used to extract the high-level feature of speech signal in each frame. The transformer and long short-Term memory (LSTM) are combined to model the sequential information along the time dimension. Furthermore, using the self-Attention mechanism, the transformer is used to compute the similarities between different positions in the long sequence and dynamically generate the weights for different positions. Thus, the lost packets can be better predicted by explicitly providing attention weights for the previous frames. The experimental results show that the proposed method has better performance in perceptual evaluation of speech quality (PESQ) and short-Term objective intelligibility (STOI). © 2023 IEEE.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：