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： Wu, Ning-Qian Hu, Ya-Jun Chen, Liping Ling, Zhen-Hua National Engineering Research Center of Speech and Language Information Processing University of Science and Technology of China Hefei China iFLYTEK Research iFLYTEK Co. Ltd. China MoE Key Laboratory of Brain-inspired Intelligent Perception and Cognition University of Science and Technology of China China

ISBN: (纸本)9798350368741

Spontaneous behaviors in speech pose significant challenges for speech synthesis. Existing research has not adequately addressed these behaviors, with most studies relying on specially recorded datasets. In contrast, real-world data more accurately reflects the natural, spontaneous speaking styles in everyday life and encompasses a wider range of spontaneous behaviors. However, such data is often of lower quality, and the distribution of spontaneous behaviors is highly imbalanced. In this study, we explore spontaneous speech synthesis using real-world data within the VITS2 framework. To overcome these challenges, we introduce two techniques: anchored monotonic alignment and spontaneous hidden representation substitution. Experimental results demonstrate that these methods enhance model alignment and improve the naturalness of the generated speech. Our proposed approach successfully addresses the challenge of synthesizing rare spontaneous behaviors and offers users flexible control over the synthesized speech. © 2025 IEEE.

关键词： real-world data speech synthesis spontaneous behaviors spontaneous speech synthesis VITS

来源：评论

学校读者我要写书评

暂无评论

An Unsupervised Image Enhancement Method Based on Adaptation Region Divisions

引用

IET Image processing 2025年第1期19卷

作者： Zhou, Kaijun Yuan, Weiyi Qin, Yemei School of Intelligent Engineering and Intelligent Manufacturing Hunan University of Technology and Business Changsha China Provincial Engineering Research Center for Intelligent Perception and Information Processing Changsha China

This paper presents a novel image enhancement method that integrates traditional image processing techniques with deep learning frameworks. Initially, images are transformed from the red, green and blue (RGB) color space to the Lab color space, and the luminance component (L) is extracted to quantify texture. Subsequently, texture complexity is assessed using features derived from the gray-level co-occurrence matrix (GLCM), including contrast, correlation, homogeneity, and energy. These features are weighted to compute an overall texture complexity score, which facilitates the segmentation of the image into distinct regions. Regions characterized by simple textures are aggregated into larger segments, whereas regions with complex textures are subdivided into smaller segments. Following segmentation, histogram equalization is applied along with noise reduction and image enhancement via a convolutional autoencoder model. The model extracts relevant features and reduces dimensionality in the encoder phase, and reconstructs the image through the decoder. This methodology effectively preserves semantic information and enhances image clarity. Some experiments are conducted using the ExDark dataset, which comprises twelve categories, and the enhancement results are quantitatively evaluated using image quality metrics such as peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), learned perceptual image patch similarity (LPIPS), and neural image quality evaluator (NIQE). Experimental results demonstrate that the proposed method significantly surpasses existing enhancement techniques in terms of image quality and visual perception, thereby affirming its efficacy in improving the visual quality and detail of low-light images. The implementation code will be made publicly available at: https://***/Winnie0320/Image-Enhancement-Method. © 2025 The Author(s). IET Image processing published by John Wiley & Sons Ltd on behalf of The Institution of Engineering

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

Geometry-Aware 3D Salient Object Detection Network

arXiv

引用

arXiv 2025年

作者： Wang, Chen Zhang, Liyuan Hui, Le Liu, Qi Dai, Yuchao Shaanxi Key Laboratory of Information Acquisition and Processing Northwestern Polytechnical University China PCA Lab Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education Nanjing University of Science and Technology China

Point cloud salient object detection has attracted the attention of researchers in recent years. Since existing works do not fully utilize the geometry context of 3D objects, blurry boundaries are generated when segmenting objects with complex backgrounds. In this paper, we propose a geometry-aware 3D salient object detection network that explicitly clusters points into superpoints to enhance the geometric boundaries of objects, thereby segmenting complete objects with clear boundaries. Specifically, we first propose a simple yet effective superpoint partition module to cluster points into superpoints. In order to improve the quality of superpoints, we present a point cloud class-agnostic loss to learn discriminative point features for clustering superpoints from the object. After obtaining superpoints, we then propose a geometry enhancement module that utilizes superpoint-point attention to aggregate geometric information into point features for predicting the salient map of the object with clear boundaries. Extensive experiments show that our method achieves new state-of-the-art performance on the PCSOD dataset. Copyright © 2025, The Authors. All rights reserved.

关键词： Geometry

来源：评论

学校读者我要写书评

暂无评论

Learning to Learn Transferable Generative Attack for Person Re-Identification

引用

IEEE Transactions on Image processing 2025年 PP卷 PP页

作者： Bian, Yuan Liu, Min Wang, Xueping Ma, Yunfeng Wang, Yaonan Hunan University National Engineering Research Center of Robot Visual Perception and Control Technology College of Electrical and Information Engineering Hunan Changsha China Hunan Normal University Hunan Provincial Key Laboratory of Intelligent Computing and Language Information Processing College of Information Science and Engineering Hunan Changsha China

Deep learning-based person re-identification (reid) models are widely employed in surveillance systems and inevitably inherit the vulnerability of deep networks to adversarial attacks. Existing attacks merely consider cross-dataset and cross-model transferability, ignoring the cross-test capability to perturb models trained in different domains. To powerfully examine the robustness of real-world re-id models, the Meta Transferable Generative Attack (MTGA) method is proposed, which adopts meta-learning optimization to promote the generative attacker producing highly transferable adversarial examples by learning comprehensively simulated transfer-based crossmodel&dataset&test black-box meta attack tasks. Specifically, cross-model&dataset black-box attack tasks are first mimicked by selecting different re-id models and datasets for meta-train and meta-test attack processes. As different models may focus on different feature regions, the Perturbation Random Erasing module is further devised to prevent the attacker from learning to only corrupt model-specific features. To boost the attacker learning to possess cross-test transferability, the Normalization Mix strategy is introduced to imitate diverse feature embedding spaces by mixing multi-domain statistics of target models. Extensive experiments show the superiority of MTGA, especially in cross-model&dataset and cross-model&dataset&test attacks, our MTGA outperforms the SOTA methods by 20.0% and 11.3% on mean mAP drop rate, respectively. © 1992-2012 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Colocalized Raman and IR Spectroscopies via Vibrational-Encoded Fluorescence for Comprehensive Vibrational Analysis

引用

Journal of the American Chemical Society 2025年第19期147卷 16309-16318页

作者： Meng, Zhao-Dong Wu, Tai-Rui Zhou, Li-Ling You, En-Ming Dong, Zhi-Peng Zhang, Xia-Guang Chen, Gan-Yu Wu, De-Yin Yi, Jun Tian, Zhong-Qun School of Electronic Science and Engineering State Key Laboratory of Physical Chemistry of Solid Surfaces College of Chemistry and Chemical Engineering Fujian Key Laboratory of Ultrafast Laser Technology and Applications IKKEM Xiamen University Xiamen361005 China School of Ocean Information Engineering Fujian Provincial Key Laboratory of Oceanic Information Perception and Intelligent Processing Jimei University Xiamen361021 China Key Laboratory of Green Chemical Media and Reactions Ministry of Education School of Chemistry and Chemical Engineering Henan Normal University Xinxiang453007 China

Vibrational spectroscopy, including Raman scattering and infrared (IR) absorption, provides essential molecular fingerprint information, facilitating diverse applications, such as interfacial sensing, chemical analysis, and biomedical diagnostics. The complementary selection rules of Raman and IR spectroscopies offer distinct, yet mutually reinforcing, insights into molecular structure and dynamics. However, in dynamic or complex chemical environments, either technique alone is not capable of providing a complete and nuanced picture of molecular vibrations. Simultaneous detection of complementary vibrational modes within the same molecular group remains challenging due to wavelength discrepancies and sensitivity mismatches between Raman and IR spectroscopies. In this work, to address the gap between these spectroscopies, we developed an integrated approach based on vibrational-encoded fluorescence (VEF), in which the complementary vibrational information is respectively encoded into the different parts of fluorescence radiation: Stokes fluorescence carrying Raman information and anti-Stokes fluorescence reflecting IR information. This method employs a dual-resonant microsphere-on-mirror plasmonic structure to bridge the waveband gap, enabling the simultaneous detection of complete vibrational modes in the visible spectrum with ultrahigh sensitivity down to ∼100 molecules. Hyperspectral colocalization imaging demonstrates spatial correlations between the complementary vibrations. By careful calibration, the detection efficiency is improved by 8 orders of magnitude compared to unenhanced IR spectroscopy. This approach creates new opportunities for the precise identification of molecular vibrational information in complex chemical environments. © 2025 American Chemical Society.

关键词： Vibrational spectroscopy

来源：评论

学校读者我要写书评

暂无评论

IsoVISoR: Towards 3D Mesoscale Brain Mapping of Large Mammals at Isotropic Sub-micron Resolution

引用

Neuroscience Bulletin 2025年第2期41卷 344-348页

作者： Chao-Yu Yang Yan Shen Xiaoyang Qi Lufeng Ding Yanyang Xiao Qingyuan Zhu Hao Wang Cheng Xu Pak-Ming Lau Pengcheng Zhou Fang Xu Guo-Qiang Bi Division of Life Sciences and Medicine CAS Key Laboratory of Brain Function and DiseaseUniversity of Science and Technology of ChinaHefei230026China Interdisciplinary Center for Brain Information Shenzhen-Hong Kong Institute of Brain ScienceShenzhen Institute of Advanced TechnologyChinese Academy of SciencesShenzhen518055China Center for Integrative Imaging Hefei National Research Center for Physical Sciences at the MicroscaleUniversity of Science and Technology of ChinaHefei230027China National Engineering Laboratory for Brain-Inspired Intelligence Technology and Application MoE Key Laboratory of Brain-Inspired Intelligent Perception and CognitionSchool of Information Science and TechnologyUniversity of Science and Technology of ChinaHefei230026China Anhui Province Key Laboratory of Biomedical Imaging and Intelligent Processing Institute of Artificial IntelligenceHefei Comprehensive National Science CenterHefei230088China Faculty of Life and Health Sciences Shenzhen University of Advanced TechnologyShenzhen518107China CAS Key Laboratory of Brain Connectome and Manipulation The Brain Cognition and Brain Disease InstituteShenzhen Institute of Advanced TechnologyChinese Academy of SciencesShenzhen518055China

Dear Editor,The mammalian brain exhibits cross-scale complexity in neuronal morphology and connectivity,the study of which demands high-resolution morphological reconstruction of individual neurons across the entire brain[1-4].Current commonly used approaches for such mesoscale brain mapping include two main types of three-dimensional fluorescence microscopy:the block-face methods,and the lightsheet-based methods[5,6].In general,the high imaging speed and light efficiency of light-sheet microscopy make it a suitable tool for high-throughput volumetric imaging,especially when combined with tissue-clearing ***,large brain samples pose major challenges to this approach.

关键词： brain mapping three dimensional fluorescence microscopy mesoscale brain mapping lightsheet based methods block face methods high throughput volumetric imaging tissue clearing techniques

来源：评论

学校读者我要写书评

暂无评论

Modality Unified Attack for Omni-Modality Person Re-Identification

引用

IEEE Transactions on information Forensics and Security 2025年

作者： Bian, Yuan Liu, Min Yi, Yunqi Wang, Xueping Ma, Yunfeng Wang, Yaonan Hunan University National Engineering Research Center of Robot Visual Perception and Control Technology College of Electrical and Information Engineering Hunan Changsha China Hunan Normal University Hunan Provincial Key Laboratory of Intelligent Computing and Language Information Processing College of Information Science and Engineering Hunan Changsha China

Deep learning based person re-identification (re-id) models have been widely employed in surveillance systems. Recent studies have demonstrated that black-box single-modality and cross-modality re-id models are vulnerable to adversarial examples (AEs), leaving the robustness of multi-modality re-id models unexplored. Due to the lack of knowledge about the specific type of model deployed in the target black-box surveillance system, we aim to generate modality unified AEs for omni-modality (single-, cross- and multi-modality) re-id models. Specifically, we propose a novel Modality Unified Attack method to train modality-specific adversarial generators to generate AEs that effectively attack different omni-modality models. A multi-modality model is adopted as the surrogate model, wherein the features of each modality are perturbed by metric disruption loss before fusion. To collapse the common features of omnimodality models, Cross Modality Simulated Disruption approach is introduced to mimic the cross-modality feature embeddings by intentionally feeding images to non-corresponding modality-specific subnetworks of the surrogate model. Moreover, Multi Modality Collaborative Disruption strategy is devised to facilitate the attacker to comprehensively corrupt the informative content of person images by leveraging a multi modality feature collaborative metric disruption loss. Extensive experiments show that our MUA method can effectively attack the omni-modality re-id models, achieving 55.9%, 24.4%, 49.0% and 62.7% mean mAP Drop Rate, respectively. © 2005-2012 IEEE.

关键词： Modality unified attack Multi Modality Collaborative Disruption Omni-Modality re-id

来源：评论

学校读者我要写书评

暂无评论

Capacity Maximization for Double Passive/Active RIS-Aided MISO Systems in Multi-User Scenarios

引用

IEEE Internet of Things Journal 2025年

作者： Peng, Xin Liu, Chenyu Yu, Chao Xie, Wenwu Yang, Liang Hunan Institute of Science and Technology School of Information Science and Engineering Yueyang414006 China Hunan Engineering Research Center for Autonomous Data Perception and Processing in Industrial Intelligent Networking Yueyang414006 China Hunan University College of Computer Science and Electronic Engineering Changsha410082 China

This paper addresses the problem of capacity maximization in double reconfigurable intelligent surface (RIS) -aided multiple-input single-output (MISO) systems under multi-user scenarios. A system model is proposed for double RIS-aided MISO configurations, incorporating both passive RIS (PRIS) and active RIS (ARIS). The primary objective is to maximize the minimum channel capacity across all users, and a corresponding optimization problem is formulated to achieve this goal. To overcome the challenges posed by the fractional structure of the non-convex objective function, auxiliary variables are introduced, and the bisection method is utilized. Given the highly coupled and nonconvex nature of the optimization variables, a joint alternating optimization (AO) algorithm is designed based on the bisection method and successive convex approximation (SCA), enabling iterative determination of the beamforming vectors at the base station and the reflection coefficients of RISs. The results indicate that deploying double RISs substantially enhances the system’s channel capacity through reflected links. Notably, the ARIS protocols yield higher performance gains, particularly under scenarios involving high total power and large amplification factors. In multi-user environments, the double RIS configuration significantly outperforms the single RIS setup due to the increased degrees of freedom, especially when supported by ARIS protocols. These findings underscore the substantial potential and advantages of double RIS configurations in complex communication environments. © 2014 IEEE.

关键词： Convex optimization

来源：评论

学校读者我要写书评

暂无评论

Real-Time Student Behavior Analysis via YOLOv5 with Coordinate Attention 9th

Real-Time Student Behavior Analysis via YOLOv5 with Coordina...

引用

9th International Symposium on Artificial Intelligence and Robotics, ISAIR 2024

作者： Ma, Shuai Chang, Heyou Yang, Xinru Chen, Quan Han, Jian Key Laboratory of Intelligent Information Processing Nanjing XiaoZhuang University Nanjing China School of Information and Engineering Nanjing XiaoZhuang University Nanjing China School of Computer Science Nanjing University of Information Science and Technology Nanjing China Jiangsu Key Laboratory of Image and Video Understanding for Social Safety Nanjing University of Science and Technology Nanjing China Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education Nanjing University of Science and Technology Nanjing China

ISBN: (纸本)9789819629138

The detection and recognition of student behavior play a pivotal role in the context of smart classrooms. However, conventional methods often encounter performance degradation due to challenges such as occlusion, data diversity, and complexity. To effectively addresses these issues, we propose a method that integrates YOLOv5 with coordinate attention mechanism and GhostConv for analyzing students' behavior. Specifically, the coordinate attention module enhances the localization ability of the proposed method by addressing problems related to target scale variations and uneven distribution. Additionally, we employ GhostConv to replace conventional convolutional layers in YOLOv5, reducing parameters while improving computational efficiency. To verify the validity of our method, a self-built challenging dataset for real teaching scenarios and Citypersons dataset are selected for the experiment. In comparison to the original YOLOv5, our method has achieved a 14.8% and 5.7% increase in mean average precision rates, with a 19.0% improvement in frames per second. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Anchored Monotonic Alignment and Representation Substitution for Rare Spontaneous Behaviors in Spontaneous Speech Synthesis

Anchored Monotonic Alignment and Representation Substitution...

引用

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

作者： Ning-Qian Wu Ya-Jun Hu Liping Chen Zhen-Hua Ling National Engineering Research Center of Speech and Language Information Processing University of Science and Technology of China Hefei P.R.China iFLYTEK Research iFLYTEK Co. Ltd. China MoE Key Laboratory of Brain-Inspired Intelligent Perception and Cognition University of Science and Technology of China

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

关键词： Training Accuracy Large language models Speech enhancement Signal processing Acoustics Speech synthesis

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：