检索结果-内蒙古大学图书馆

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Zhang, Leying Zhang, Wangyou Chen, Zhengyang Qian, Yanmin Auditory Cognition and Computational Acoustics Lab MoE Key Lab of Artificial Intelligence AI Institute Department of Computer Science and Engineering Shanghai Jiao Tong University Shanghai China

ISBN: (纸本)9798350368741

The acoustic background plays a crucial role in natural conversation. It provides context and helps listeners understand the environment, but a strong background makes it difficult for listeners to understand spoken words. The appropriate handling of these backgrounds is situation-dependent: Although it may be necessary to remove background to ensure speech clarity, preserving the background is sometimes crucial to maintaining the contextual integrity of the speech. Despite recent advancements in zero-shot Text-to-Speech technologies, current systems often struggle with speech prompts containing backgrounds. To address these challenges, we propose a Controllable Masked Speech Prediction strategy coupled with a dual-speaker encoder, utilizing a task-related control signal to guide the prediction of dual background removal and preservation targets. Experimental results demonstrate that our approach enables precise control over the removal or preservation of background across various acoustic conditions and exhibits strong generalization capabilities in unseen scenarios. © 2025 IEEE.

关键词： background preservation background removal flow-matching text-to-speech

来源：评论

学校读者我要写书评

暂无评论

Advancing Non-intrusive Suppression on Enhancement Distortion for Noise Robust ASR

Advancing Non-intrusive Suppression on Enhancement Distortio...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Wang, Wei Zhao, Siyi Qian, Yanmin Auditory Cognition and Computational Acoustics Lab MoE Key Lab of Artificial Intelligence AI Institute Department of Computer Science and Engineering Shanghai Jiao Tong University Shanghai China

ISBN: (纸本)9798350368741

Recent advancements in speech enhancement (SE) techniques have greatly improved speech clarity and intelligibility in challenging acoustic environments. However, integrating SE into automatic speech recognition (ASR) systems often results in performance degradation due to artifacts introduced during the enhancement process. While various methods have enhanced recognition accuracy in SE-ASR systems, they often require fine-tuning or re-training of SE or ASR models, which is impractical in many real-world applications. In this paper, we propose a lightweight distortion suppression (DS) network that addresses these artifacts without modifying the SE or ASR models, treating them as fixed black boxes. The DS module operates on the time-frequency (T-F) bands of the original and enhanced complex spectrograms, efficiently compensating for SE distortions using the original T-F information. We validate our approach through experiments on both Mandarin and English ASR tasks using monaural and multi-channel SE frontends, across various ASR backends. Results show that the DS module significantly improves the performance of SE-ASR systems, even when used with robust commercial ASR backends. © 2025 IEEE.

关键词： non-intrusive robust speech recognition speech distortion speech enhancement

来源：评论

学校读者我要写书评

暂无评论

Interface Preservation in Modeling Compressible Two-Phase Flows using Six- and Seven-Equation Formulations

Interface Preservation in Modeling Compressible Two-Phase Fl...

引用

AIAA science and Technology Forum and Exposition, AIAA SciTech Forum 2025

作者： Hatashita, Luis H. Jain, Suhas S. Daniel Guggenheim School of Aerospace Engineering Flow Physics and Computational Science Lab Georgia Institute of Technology 270 Ferst Drive NW Atlanta30332 Georgia George W. Woodruff School of Mechanical Engineering Flow Physics and Computational Science Lab Georgia Institute of Technology 801 Ferst Drive NW Atlanta30332 Georgia

ISBN: (数字)9781624107238

ISBN: (纸本)9781624107238

Classical diffuse interface methods for non-equilibrium formulations of two-phase flows are not capable of preserving interface-thickness. Furthermore, when coupled with shockcapturing schemes, the interface may diffuse in its entirety. Thus, it is of utmost importance for accurate simulation of two-phase flows to describe interface-preserving formulations. Herein, a conservative interface-regularization formulation based on the phase-field method is proposed, and it analytically guarantees boundedness of volume fraction, absence of spurious oscillations of phasic kinetic energy and mixture entropy preservation in the absence of relaxation. The formulation is shown to stabilize dispersion errors originated due to the interface jumps and to preserve interface thickness when discretized with low lost and low dissipation second-order schemes. A possible issue with this formulation and discretization is the possibility of failing to preserve positivity of quantities of interests, hence, a blending strategy with a Lax-Friedrichs flux is also proposed and tested. In highly compressible scenarios such as shock-drop interaction, the interface regularization when applied to the positive-preserving flux is demonstrated to improve qualitative results on coarser grids. © 2025, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved.

关键词： Kinetic energy

来源：评论

学校读者我要写书评

暂无评论

FEniCSx-Based Effective Scheme for the Cahn-Hilliard Equation 5

FEniCSx-Based Effective Scheme for the Cahn-Hilliard Equatio...

引用

5th IEEE International Conference on Advances in Electrical, Computing, Communication and Sustainable Technologies, ICAECT 2025

作者： Bhutra, Harsh Thomas, Bettina Ann Shri, K. Shreya Smitha, T.V. Neelima, Nizampatnam Amrita School of Engineering Amrita Vishwa Vidyapeetham Department of Electronics and Communication Bengaluru India Amrita School of Engineering Amrita Vishwa Vidyapeetham Computational Science Lab Bengaluru India

ISBN: (纸本)9798350356236

The Cahn-Hilliard equation is a fourth-order parabolic partial differential equation that is challenging and time-intensive to solve using conventional methods. Although the equation is difficult to solve, it finds its use cases in many aspects of science including but not limited to the study of phase separation in the binary, isotropic and isothermal mixtures of materials such as molten alloys. Some of the other applications are Bioinformatics, computational Biology, Biomedical informatics, human activity recognition, Healthcare, computer vision, Climate science and Natural Language Processing. Many researchers have come up with a way to deal with partial differential equations, but very few works and methods deal with different orders of partial differential equations. In this article, a novel and effective scheme for solving the Cahn-Hilliard equation is proposed using FEniCS. The time of execution and relative tolerance of the solutions of this equation are computed from methods of direct computation and parallel processing which is carried out for the equation with higher-order finite element method for various boundary conditions to demonstrate the efficiency of the proposed approach. The solution thus obtained helps in understanding the non-linear behaviour of the equation with the help of a three-dimensional heatmap plot. This work solves the Cahn Hilliard equation by using the in-built libraries, Dolphin an open-source software called FEniCS and Message Passing Interface (MPI). The MPI is responsible for achieving parallelism in high-performance computing environments and breaks down the problem into four smaller parts making it faster to get to the solution. The obtained results demonstrate that this approach outperforms the existing literature by achieving a forty percent faster computational speed. The approach will be very useful in solving many similar problems arising in the field of science and engineering. © 2025 IEEE.

关键词： Open source software

来源：评论

学校读者我要写书评

暂无评论

Advanced Flow Dynamics Mapping Through Volumetric Photoacoustic Particle Velocimetry 17

Advanced Flow Dynamics Mapping Through Volumetric Photoacous...

引用

17th International Conference on Photonics and Imaging in Biology and Medicine, PIBM 2024

作者： Gao, Xiali Xiao, Zhibo Wang, Xuanhao Shi, Junhui College of Biomedical Engineering and Instrument Science Zhejiang University Hangzhou310027 China Research Center for Novel Computational Sensing and Intelligent Processing Zhejiang Lab Hangzhou311100 China

ISBN: (数字)9781510688070

ISBN: (纸本)9781510688063

Flow field analysis is vital for biomedical applications such as drug delivery, dosage optimization, and hemodynamic studies. Conventional optical particle image velocimetry (PIV) struggles with dynamic three-dimensional analysis in turbid media. This study introduces a photoacoustic volumetric particle image velocimetry (PAV-PIV) system to address these challenges, enabling rapid three-dimensional flow field reconstruction in optically opaque environments without complex calibration. Vascular phantoms made from silicone, polyurethane, and agarose were tested, with the optimal material selected based on acoustic impedance contrast. Nickel-coated polystyrene particles served as tracers, uniformly dispersed in the fluid. Particle displacement was measured at 20 ms intervals, allowing for flow field analysis, including velocity and vorticity distributions. Reconstructed three-dimensional photoacoustic signals demonstrated superior imaging quality in turbid media, with detailed flow field characteristics derived from the best-quality images. The system's performance and measurement precision were further evaluated. Results confirm that the PAV-PIV system effectively overcomes the limitations of traditional optical methods, offering robust and dynamic flow field analysis in challenging environments. This approach shows great potential for advancing real-time biomedical applications requiring the analysis of complex flow behaviors. © 2025 SPIE.

关键词： Drug dosage

来源：评论

学校读者我要写书评

暂无评论

Benchmarking model chemistry composite calculations for vertical ionization potential of molecular systems

引用

Chinese Chemical Letters 2025年第2期36卷 508-511页

作者： Bingwei Wang Yihong Ding Xiao Tian Key Laboratory of Carbon Materials of Zhejiang Province Wenzhou Key Lab of Advanced Energy Storage and ConversionZhejiang Province Key Lab of Leather EngineeringCollege of Chemistry and Materials EngineeringWenzhou UniversityWenzhou 325035China School of Mathematics and Science Hebei GEO UniversityShijiazhuang 050031China Laboratory of Theoretical and Computational Chemistry Institute of Theoretical ChemistryJilin UniversityChangchun 130023China

In chemical science,the vertical ionization potential(VIP)is a crucial metric for understanding the electronegativity,hardness and softness of chemical material systems as well as the electronic structure and stability of *** since the last century,the model chemistry composite methods have witnessed tremendous developments in computing the thermodynamic properties as well as the barrier ***,their performance in realm of the vertical electron processes of molecular systems has been rarely *** this study,we for the first time benchmarked the model chemistry composite methods(e.g.,CBS-QB3,G4 and W1BD)in comparison with the commonly used Koopmans's theorem(KT),electron propagator theory(e.g.,OVGF,D2,P3 and P3+)and CCSD(T)methods in calculating the VIP for up to 613 molecular systems with available experimental *** large-scale test calculations strongly showed that the CBS-QB3 model chemistry composite technique can be well recommended to calculate VIP from the perspectives of accuracy,economy and ***,the VIP values of up to 7 molecules were identified to have the absolute errors of larger than 0.3 e V at all calculation levels,which have strong hints that their VIP experimental values should be re-investigated.

关键词： Model chemistry Vertical ionization potential Composite methods Molecular systems CBS-QB3

来源：评论

学校读者我要写书评

暂无评论

Ptychographic characterization of spatiotemporal wavepackets

引用

Optica 2025年第4期12卷 498-504页

作者： Liu, Yuanyuan Xu, Yingming Zhang, Nianjia Liu, Dawei Wei, Lijin Cao, Qian Shi, Junhui Zhan, Qiwen School of Optical-Electrical and Computer Engineering University of Shanghai for Science and Technology Shanghai200093 China Zhangjiang Laboratory 100 Haike Road Shanghai201204 China Research Center for Novel Computational Sensing and Intelligent Processing Zhejiang Lab. Hangzhou311100 China College of Mechatronics and Control Engineering Shenzhen University Shenzhen518060 China

Exciting advances in spatiotemporal wavepackets have shown great potential in a range of applications, including particle manipulation, time-resolved imaging, and unique light–matter interactions. In particular, spatiotemporal optical vortice (STOV) pulse beams characterized by a helical phase structure in space and time are a recently emerging type of complex structured ultrashort pulse. However, the high-resolution measurement and manipulation of ultrashort spatiotemporal wavepackets still face major challenges due to their inseparable properties in space and time. In this study, based on spectra-encoded ptychography, we implement a simple, compact, reference-free STOV characterization technique with a temporal resolution of 6 fs and a spatial resolution of 11.05 µm. This technique can be employed to analyze various intricate spatiotemporal structural beams, contributing to the growing field of spatiotemporally engineered ultrafast laser pulse beams. Additionally, as a lensless measurement technique, this characterization method may be extended to other spectral ranges, including the extreme ultraviolet and X-ray domains. © 2025 Optica Publishing Group.

关键词： Wave packets

来源：评论

学校读者我要写书评

暂无评论

Generalizable Audio Deepfake Detection via Latent Space Refinement and Augmentation

Generalizable Audio Deepfake Detection via Latent Space Refi...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Huang, Wen Gu, Yanmei Wang, Zhiming Zhu, Huijia Qian, Yanmin Auditory Cognition and Computational Acoustics Lab MoE Key Lab of Artificial Intelligence AI Institute Department of Computer Science and Engineering Shanghai Jiao Tong University Shanghai China SJTU Paris Elite Institute of Technology Shanghai China Ant Group Shanghai China

ISBN: (纸本)9798350368741

Advances in speech synthesis technologies, like text-to-speech (TTS) and voice conversion (VC), have made detecting deepfake speech increasingly challenging. Spoofing countermeasures often struggle to generalize effectively, particularly when faced with unseen attacks. To address this, we propose a novel strategy that integrates Latent Space Refinement (LSR) and Latent Space Augmentation (LSA) to improve the generalization of deepfake detection systems. LSR introduces multiple learnable prototypes for the spoof class, refining the latent space to better capture the intricate variations within spoofed data. LSA further diversifies spoofed data representations by applying augmentation techniques directly in the latent space, enabling the model to learn a broader range of spoofing patterns. We evaluated our approach on four representative datasets, i.e. ASVspoof 2019 LA, ASVspoof 2021 LA and DF, and In-The-Wild. The results show that LSR and LSA perform well individually, and their integration achieves competitive results, matching or surpassing current state-of-the-art methods. © 2025 IEEE.

关键词： anti-spoofing audio deepfake detection generalization

来源：评论

学校读者我要写书评

暂无评论

A Role of Wearable Health Technology in Smart Cities 3rd

A Role of Wearable Health Technology in Smart Cities

引用

3rd International Conference on computational Electronics for Wireless Communications, ICCWC 2023

作者： Chauhan, Ritu Kaur, Harleen Mehta, Khushi Alankar, Bhavya Artificial Intelligence and IoT Lab Center for Computational Biology and Bioinformatics Amity University Uttar Pradesh Noida India Department of Computer Science and Engineering School of Engineering Sciences and Technology Jamia Hamdard New Delhi India Amity Institute of Biotechnology Amity University Uttar Pradesh Noida India

ISBN: (纸本)9789819719426

In this rapidly changing era of the world, healthcare devices are revolutionizing, which enables the continuous measurement of crucial biological markers for diagnostic purposes, physiological health monitoring, and evaluation. Hence, wearable devices have transformed into accessories, embedded clothing, body adaptors, and body implants. Additionally, the emergence of wearable healthcare devices that facilitates the diagnosis and prognosis by using tiny sensing devices and biomedical devices has greatly enhanced the efficiency as well as the quality of medical care facilities results in the tremendous advancements in semiconductor technology, biomedical technologies, and nano-sized materials over the past few decades. Meanwhile, the devices could provide real-time feedback, allowing users to analyze their health and monitor in a specific period that has been assigned to the smart wearable devices, even though the industries and manufacturers are intensely interested in interpreting the factors that influence the adoption of new smart technologies, which helps to improve the performance and viability of wearable devices to fascinate consumers. Hence, the advancements in technology are regarded as trustworthy equipment for everlasting monitoring of health that are frequently used to monitor and control a diverse of health tracking predictors in the surroundings, including monitoring health and wellness of the body. In recent years, it has been focused on the study of a wide range of parameters closer to advancements in smart technology. As a result, these smart devices are now utilized in a variety of healthcare monitoring applications, one of the most crucial aspects of data collection is the various factors based on bodily activities. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

Artificial Intelligence as Automated Technology for Prediction of Breast Cancer 3rd

Artificial Intelligence as Automated Technology for Predicti...

引用

3rd International Conference on computational Electronics for Wireless Communications, ICCWC 2023

作者： Chauhan, Ritu Kaur, Harleen Choudhary, Tisya Alankar, Bhavya Artificial Intelligence and IoT Lab Center for Computational Biology and Bioinformatics Amity University Uttar Pradesh Noida India Department of Computer Science and Engineering School of Engineering Sciences and Technology Jamia Hamdard New Delhi India Amity Institute of Biotechnology Amity University Uttar Pradesh Noida India

ISBN: (纸本)9789819719426

Artificial intelligence techniques are utilized in cancer research and oncology which involves medical data in the detection of cancer, subtype classification, and optimization of cancer treatment. This paper discusses artificial intelligence techniques, such as K-means cluster analysis and its data interpretation based on breast cancer diagnostic dataset. This study focuses on the database of digital images of fine needle aspirate (FNA) of breast mass which shows abnormal lump, and a clustering technique was applied on this dataset. This study represents various descriptive statistics techniques of the breast cancer diagnostic dataset. Furthermore, we have also tried to show a variety of techniques of descriptive statistics like KMO and Bartlett’s test, total variance representing initial eigenvalues and extraction sum of squared loadings, scree plots, and clustering techniques like ANOVA table, initial and final cluster centers, and the distance between initial and final cluster centers of K-means cluster analysis. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Lung cancer

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：