检索结果-内蒙古大学图书馆

48th IEEE International Conference on Acoustics, Speech and signal processing, ICASSP 2023

作者： Yu, Jianwei Chen, Hangting Luo, Yi Gu, Rongzhi Li, Weihua Weng, Chao Audio and Speech Signal Processing Oteam Tencent Ai Lab China

ISBN: (纸本)9781728163277

This report presents the development of Tencent AI lab's personalized speech enhancement system for the 2023 ICASSP signal processing Grand Challenge - deep noise suppression (DNS) challenge1, which includes the use of a modified band-split recurrent neural network (BSRNN) and a multi-resolution spectrogram discriminator to improve perceptual quality metrics. The proposed system outperforms the baseline system by 0.047 and 0.036 final scores in the headset track and speaker phone track, respectively, and ranks among the top-32 in both tracks. © 2023 IEEE.

关键词： Speech enhancement

来源：评论

学校读者我要写书评

暂无评论

Complexity Scaling for Speech Denoising

Complexity Scaling for Speech Denoising

引用

International Conference on Acoustics, Speech, and signal processing (ICASSP)

作者： Hangting Chen Jianwei Yu Chao Weng Tencent AI Lab Audio and Speech Signal Processing Oteam

Computational complexity is critical when deploying deep learning-based speech denoising models for on-device applications. Most prior research focused on optimizing model architectures to meet specific computational cost constraints, often creating distinct neural network architectures for different complexity limitations. This study conducts complexity scaling for speech denoising tasks, aiming to consolidate models with various complexities into a unified architecture. We present a Multi-Path Transform-based (MPT) architecture to handle both low- and high-complexity scenarios. A series of MPT networks present high performance covering a wide range of computational complexities on the DNS challenge dataset. Moreover, inspired by the scaling experiments in natural language processing, we explore the empirical relationship between model performance and computational cost on the denoising task. As the complexity number of multiply-accumulate operations (MACs) is scaled from 50M/s to 15G/s on MPT networks, we observe a linear increase in the values of PESQ-WB and SI-SNR, proportional to the logarithm of MACs, which might contribute to the understanding and application of complexity scaling in speech denoising tasks. 1

关键词：

来源：评论

学校读者我要写书评

暂无评论

Neural Optimization of Geometry and Fixed Beamformer for Linear Microphone Arrays 48

Neural Optimization of Geometry and Fixed Beamformer for Lin...

引用

48th IEEE International Conference on Acoustics, Speech and signal processing, ICASSP 2023

作者： Yan, Longfei Huang, Weilong Kleijn, W. Bastiaan Abhayapala, Thushara D. Victoria University of Wellington School of Engineering and Computer Science New Zealand Australian National University Audio & Acoustic Signal Processing Group Australia Alibaba Group DingTalk Hummingbird Audio Lab China

ISBN: (纸本)9781728163277

Fixed beamforming based on uniform linear microphone arrays often suffers from non-optimal performance for broadband signals. This paper addresses the issue by jointly optimizing the array geometry and spatial filters through a neural network based model. The model, composed of two feed forward neural networks, is optimized in an end-to-end manner. It satisfies the distortionless constraint in the look direction. Experimental results show that the proposed model outperforms the previous state-of-the-art fixed beamformer with overall better scores. Moreover, the proposed model can control the tradeoff between Directivity Factor (DF) and White Noise Gain (WNG) in a flexible way. © 2023 IEEE.

关键词： Feedforward neural networks

来源：评论

学校读者我要写书评

暂无评论

audioComposer: Towards Fine-grained audio Generation with Natural Language Descriptions

AudioComposer: Towards Fine-grained Audio Generation with Na...

引用

2025 IEEE International Conference on Acoustics, Speech, and signal processing, ICASSP 2025

作者： Wang, Yuanyuan Chen, Hangting Yang, Dongchao Wu, Zhiyong Wu, Xixin The Chinese University of Hong Kong Hong Kong Tencent AI Lab Audio and Speech Signal Processing Oteam China Shenzhen International Graduate School Tsinghua University Shenzhen China

ISBN: (纸本)9798350368741

Current Text-to-audio (TTA) models mainly use coarse text descriptions as inputs to generate audio, which hinders models from generating audio with fine-grained control of content and style. Some studies try to improve the granularity by incorporating additional frame-level conditions or control networks. However, this usually leads to complex system design and difficulties due to the requirement for reference frame-level conditions. To address these challenges, we propose audioComposer, a novel TTA generation framework that relies solely on natural language descriptions (NLDs) to provide both content specification and style control information. To further enhance audio generative modeling, we employ flow-based diffusion transformers with the cross-attention mechanism to incorporate text descriptions effectively into audio generation processes, which can not only simultaneously consider the content and style information in the text inputs, but also accelerate generation compared to other architectures. Furthermore, we propose a novel and comprehensive automatic data simulation pipeline to construct data with fine-grained text descriptions, which significantly alleviates the problem of data scarcity in the area. Experiments demonstrate the effectiveness of our framework using solely NLDs as inputs for content specification and style control. The generation quality and controllability surpass state-of-the-art TTA models, even with a smaller model size. © 2025 IEEE.

关键词： audio generation diffusion flow-based natural language description style control

来源：评论

学校读者我要写书评

暂无评论

TSpeech-AI System Description to the 5th Deep Noise Suppression (DNS) Challenge

TSpeech-AI System Description to the 5th Deep Noise Suppress...

引用

International Conference on Acoustics, Speech, and signal processing (ICASSP)

作者： Jianwei Yu Hangting Chen Yi Luo Rongzhi Gu Weihua Li Chao Weng Tencent AI Lab Audio and Speech Signal Processing Oteam

This report presents the development of Tencent AI lab’s personalized speech enhancement system for the 2023 ICASSP signal processing Grand Challenge – deep noise suppression (DNS) challenge 1 , which includes the use of a modified band-split recurrent neural network (BSRNN) and a multi-resolution spectrogram discriminator to improve perceptual quality metrics. The proposed system outperforms the baseline system by 0.047 and 0.036 final scores in the headset track and speaker phone track, respectively, and ranks among the top-3 2 in both tracks.

关键词： Measurement Headphones Recurrent neural networks Noise reduction Speech enhancement Acoustics Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Ultra Dual-Path Compression For Joint Echo Cancellation And Noise Suppression

arXiv

引用

arXiv 2023年

作者： Chen, Hangting Yu, Jianwei Luo, Yi Gu, Rongzhi Li, Weihua Lu, Zhuocheng Weng, Chao Tencent AI Lab Audio and Speech Signal Processing Oteam

Echo cancellation and noise reduction are essential for full-duplex communication, yet most existing neural networks have high computational costs and are inflexible in tuning model complexity. In this paper, we introduce time-frequency dual-path compression to achieve a wide range of compression ratios on computational cost. Specifically, for frequency compression, trainable filters are used to replace manually designed filters for dimension reduction. For time compression, only using frame skipped prediction causes large performance degradation, which can be alleviated by a post-processing network with full sequence modeling. We have found that under fixed compression ratios, dual-path compression combining both the time and frequency methods will give further performance improvement, covering compression ratios from 4x to 32x with little model size change. Moreover, the proposed models show competitive performance compared with fast FullSubNet and DeepFilterNet. 1 Copyright © 2023, The Authors. All rights reserved.

关键词： Echo suppression

来源：评论

学校读者我要写书评

暂无评论

COMPLEXITY SCALING FOR SPEECH DENOISING

arXiv

引用

arXiv 2023年

作者： Chen, Hangting Yu, Jianwei Weng, Chao Tencent AI Lab Audio and Speech Signal Processing Oteam

关键词： Network architecture

来源：评论

学校读者我要写书评

暂无评论

A Multi-Noise Multi-Channel ANC System using Relative Transfer Matrix-Based Approach

A Multi-Noise Multi-Channel ANC System using Relative Transf...

引用

International Workshop on Acoustic signal Enhancement (IWAENC)

作者： Yile Angela Zhang Thushara D. Abhayapala Huiyuan Sun Prasanga N. Samarasinghe Amy Bastine Audio and Acoustic Signal Processing Group The Australian National University Australia Computing and Audio Research Lab The University of Sydney NSW Australia

ISBN: (数字)9798350361858

ISBN: (纸本)9798350361865

Remote microphone-based virtual sensing (RM-VS) methods are used in active noise control (ANC) systems to avoid the physical presence of microphones inside the region of interest (ROI). Conventionally, this is achieved by estimating an observation filter to obtain the primary noise signals at virtual microphones inside the ROI, with the physical microphones recording outside the ROI. In this paper, we present a novel structure of the RM technique, where we estimate the relative transfer matrix (ReTM) between the signals at virtual microphones and physical microphones with both noise sources and secondary sources active. This ReTM can be directly applied to the recorded residual error signals during noise control, as opposed to the conventional observation filter which is typically applied to the estimated primary noise signals. Simulation results show that the proposed ReTM-based approach achieves similar performance to the conventional RM method in a stationary environment, while being more robust to changes in the secondary paths.

关键词： Soft sensors Simulation Noise Noise reduction Filtering algorithms Robustness Acoustics Recording Monitoring Microphones

来源：评论

学校读者我要写书评

暂无评论

Active Noise Control Over 3D Space with A Dynamic Noise Source

Active Noise Control Over 3D Space with A Dynamic Noise Sour...

引用

International Conference on Acoustics, Speech, and signal processing (ICASSP)

作者： Huiyuan Sun Craig T. Jin Thushara Abhayapala Prasanga Samarasinghe Audio & Acoustic Signal Processing Group The Australian National University ACT Australia Computing and Audio Research Lab The University of Sydney NSW Australia

Spatial Active noise control (ANC) systems are proposed to minimize the noise over a spatial region of interest around people’s heads by generating an anti-noise field with multiple microphones and loudspeakers. Recently, a realistic microphone geometry was designed to allow the system to monitor the residual noise field without restricting people’s head movement. However, this design, which utilizes the remote microphone technique using an observation filter (OF) for noise field modelling, has a drawback that the system is not robust against variations in the noise source location. In this paper, we address this drawback and present an improved spatial ANC system that overcomes this limitation. The proposed system enhances its adaptability by selectively employing the most suitable OF from a set of pre-modelled discriminator filter (DF), tailored to the acoustic environment of the system. We demonstrate that the proposed method maintains noise reduction performance over the region of interest with a dynamic noise source with changing position and signal through simulation and physical measurements.

关键词：

来源：评论

学校读者我要写书评

暂无评论

VIRTUAL ANALOG MODELING OF DISTORTION CIRCUITS USING NEURAL ORDINARY DIFFERENTIAL EQUATIONS 25

VIRTUAL ANALOG MODELING OF DISTORTION CIRCUITS USING NEURAL ...

引用

25th International Conference on Digital audio Effects, DAFx 2022

作者： Wilczek, Jan Wright, Alec Välimäki, Vesa Habets, Emanuël A.P. WolfSound Katowice Poland Acoustics Lab Dept. Signal Processing and Acoustics Aalto University Espoo Finland International Audio Laboratories Erlangen Erlangen Germany

ISBN: (纸本)9783200085992

Recent research in deep learning has shown that neural networks can learn differential equations governing dynamical systems. In this paper, we adapt this concept to Virtual Analog (VA) modeling to learn the ordinary differential equations (ODEs) governing the first-order and the second-order diode clipper. The proposed models achieve performance comparable to state-of-the-art recurrent neural networks (RNNs) albeit using fewer parameters. We show that this approach does not require oversampling and allows to increase the sampling rate after the training has completed, which results in increased accuracy. Using a sophisticated numerical solver allows to increase the accuracy at the cost of slower processing. ODEs learned this way do not require closed forms but are still physically interpretable. Copyright: © 2022 Jan Wilczek et al.

关键词： Ordinary differential equations

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：