检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Wang, Di Zhang, Jing Xu, Minqiang Liu, Lin Wang, Dongsheng Gao, Erzhong Han, Chengxi Guo, Haonan Du, Bo Tao, Dacheng Zhang, Liangpei School of Computer Science Wuhan University Wuhan430072 China Institute of Artificial Intelligence Wuhan University Wuhan430072 China National Engineering Research Center for Multimedia Software Wuhan University Wuhan430072 China Hubei Key Laboratory of Multimedia and Network Communication Engineering Wuhan University Wuhan430072 China School of Computer Science Faculty of Engineering The University of Sydney Australia iFlytek Co Ltd National Engineering Research Center of Speech and Language Information Processing Hefei230088 China State Key Laboratory of Information Engineering in Surveying Mapping and Remote Sensing Wuhan University Wuhan430079 China School of Computer Science and Engineering Nanyang Technological University Singapore Singapore

Foundation models have reshaped the landscape of Remote Sensing (RS) by enhancing various image interpretation tasks. Pretraining is an active research topic, encompassing supervised and self-supervised learning methods to initialize model weights effectively. However, transferring the pretrained models to downstream tasks may encounter task discrepancy due to their formulation of pretraining as image classification or object discrimination tasks. In this study, we explore the Multi-Task Pretraining (MTP) paradigm for RS foundation models to address this issue. Using a shared encoder and task-specific decoder architecture, we conduct multi-task supervised pretraining on the SAMRS dataset, encompassing semantic segmentation, instance segmentation, and rotated object detection. MTP supports both convolutional neural networks and vision transformer foundation models with over 300 million parameters. The pretrained models are finetuned on various RS downstream tasks, such as scene classification, horizontal and rotated object detection, semantic segmentation, and change detection. Extensive experiments across 14 datasets demonstrate the superiority of our models over existing ones of similar size and their competitive performance compared to larger state-of-the-art models, thus validating the effectiveness of MTP. The codes and pretrained models will be released at https://***/ViTAE-Transformer/MTP. Copyright © 2024, The Authors. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Biomedical data and AI

引用

Science China Life Sciences 2025年第5期68卷 1536-1540页

作者： Hao Xu Shibo Zhou Zefeng Zhu Vincenzo Vitelli Liangyi Chen Ziwei Dai Ning Yang Luhua Lai Shengyong Yang Sergey Ovchinnikov Zhuoran Qiao Sirui Liu Chen Song Jianfeng Pei Han Wen Jianfeng Feng Yaoyao Zhang Zhengwei Xie Yang-Yu Liu Zhiyuan Li Fulai Jin Hao Li Mohammad Lotfollahi Xuegong Zhang Ge Yang Shihua Zhang Ge Gao Pulin Li Qi Liu Jing-Dong Jackie Han Peking-Tsinghua Center for Life Sciences (CLS) Academy for Advanced Interdisciplinary StudiesPeking University Center for Quantitative Biology (CQB) Academy for Advanced Interdisciplinary StudiesPeking University Peking University-Tsinghua University-National Institute of Biological Sciences Joint Graduate Program Academy for Advanced Interdisciplinary StudiesPeking University Department of Physics University of Chicago School of Life Sciences Southern University of Science and Technology Peking University Chengdu Academy for Advanced Interdisciplinary Biotechnologies College of Chemistry and Molecular Engineering Peking University Department of Biotherapy Cancer Center and State Key Laboratory of BiotherapyWest China HospitalSichuan University Department of Biology Massachusetts Institute of Technology Lambic Therapeutics Inc. Changping Laboratory Al for Science Institute Key Laboratory of Computational Neuroscience and Brain-Inspired Intelligence Fudan University Department of Obstetrics and Gynecology West China Second University HospitalSichuan University Peking University International Cancer Institute and Peking University-Yunnan Baiyao International Medical Institute and State Key Laboratory of Natural and Biomimetic Drugs Department of Molecular and Cellular PharmacologySchool of Pharmaceutical SciencesPeking University Health Science CenterPeking University Channing Division of Network Medicine Department of MedicineBrigham and Women's Hospital and Harvard Medical School Center for Artificial Intelligence and Modeling the Carl R.Woese Institute for Genomic BiologyUniversity of Illinois Urbana-Champaign Department of Genetics and Genome Sciences School of Medicine and Department of Computer and Data Sciences and Department of Population and Quantitative Health SciencesCase Western Reserve University Department of Biochemistry and Biophysics University of California Sanger Institute Department of Automation Tsinghua University State Key Laboratory of Multimodal Artificial Intelligence Systems I

The development of artificial intelligence(AI) and the mining of biomedical data complement each other. From the direct use of computer vision results to analyze medical images for disease screening, to now integrating biological knowledge into models and even accelerating the development of new AI based on biological discoveries, the boundaries of both are constantly expanding, and their connections are becoming closer.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Selector-Enhancer: Learning Dynamic Selection of Local and Non-local Attention Operation for Speech Enhancement

arXiv

引用

arXiv 2022年

作者： Xu, Xinmeng Tu, Weiping Yang, Yuhong National Engineering Research Center for Multimedia Software School of Computer Science Wuhan University China Hubei Luojia Laboratory China Hubei Key Laboratory of Multimedia and Network Communication Engineering Wuhan University China

Attention mechanisms, such as local and non-local attention, play a fundamental role in recent deep learning based speech enhancement (SE) systems. However, natural speech contains many fast-changing and relatively brief acoustic events, therefore, capturing the most informative speech features by indiscriminately using local and non-local attention is challenged. We observe that the noise type and speech feature vary within a sequence of speech and the local and non-local operations can respectively extract different features from corrupted speech. To leverage this, we propose Selector-Enhancer, a dual-attention based convolution neural network (CNN) with a feature-filter that can dynamically select regions from low-resolution speech features and feed them to local or non-local attention operations. In particular, the proposed feature-filter is trained by using reinforcement learning (RL) with a developed difficulty-regulated reward that is related to network performance, model complexity, and "the difficulty of the SE task". The results show that our method achieves comparable or superior performance to existing approaches. In particular, Selector-Enhancer is potentially effective for real-world denoising, where the number and types of noise are varies on a single noisy mixture. © 2022, CC BY.

关键词： Speech enhancement

来源：评论

学校读者我要写书评

暂无评论

INJECTING SPATIAL INFORMATION FOR MONAURAL SPEECH ENHANCEMENT VIA KNOWLEDGE DISTILLATION

arXiv

引用

arXiv 2022年

作者： Xu, Xinmeng Tu, Weiping Yang, Yuhong National Engineering Research Center for Multimedia Software School of Computer Science Wuhan University China Hubei Key Laboratory of Multimedia and Network Communication Engineering Wuhan University China Hubei Luojia Laboratory China

Monaural speech enhancement (SE) provides a versatile and cost-effective approach to SE tasks by utilizing recordings from a single microphone. However, the monaural SE lags performance behind multi-channel SE as the monaural SE methods are unable to extract spatial information from one-channel recordings, which greatly limits their application scenarios. To address this issue, we inject spatial information into the monaural SE model and propose a knowledge distillation strategy to enable the monaural SE model to learn binaural speech features from the binaural SE model, which makes monaural SE model possible to reconstruct higher intelligibility and quality enhanced speeches under low signal-to-noise ratio (SNR) conditions. Extensive experiments show that our proposed monaural SE model by injecting spatial information via knowledge distillation achieves favorable performance against other monaural SE models with fewer parameters. © 2022, CC BY.

关键词： Cost effectiveness

来源：评论

学校读者我要写书评

暂无评论

Improving complex Reasoning over Knowledge Graph with Logic-Aware Curriculum Tuning

arXiv

引用

arXiv 2024年

作者： Xia, Tianle Ding, Liang Wan, Guojia Zhan, Yibing Du, Bo Tao, Dacheng School of Computer Science National Engineering Research Center for Multimedia Software Institute of Artificial Intelligence Hubei Key Laboratory of Multimedia and Network Communication Engineering Wuhan University China The University of Sydney Australia JD Explore Academy China College of Computing and Data Science Nanyang Technological University Singapore639798 Singapore

Answering complex queries over incomplete knowledge graphs (KGs) is a challenging task. Most previous works have focused on learning entity/relation embeddings and simulating first-order logic operators with various neural networks. However, they are bottlenecked by the inability to share world knowledge to improve logical reasoning, thus resulting in suboptimal performance. In this paper, we propose a complex reasoning schema over KG upon large language models (LLMs), containing a curriculum-based logical-aware instruction tuning framework, named LACT. Specifically, we augment the arbitrary first-order logical queries via binary tree decomposition, to stimulate the reasoning capability of LLMs. To address the difficulty gap among different types of complex queries, we design a simple and flexible logic-aware curriculum learning framework. Experiments across widely used datasets demonstrate that LACT has substantial improvements (brings an average +5.5% MRR score) over advanced methods, achieving the new state-of-the-art. Copyright © 2024, The Authors. All rights reserved.

关键词： Binary trees

来源：评论

学校读者我要写书评

暂无评论

Interpolation Method of Head-Related Transfer Functions Based on Common-Pole/Zero Modeling

引用

China Communications 2020年第10期17卷 170-182页

作者： Wei Chen Xiaochen Wang Ruimin Hu Gang Li Weiping Tu National Engineering Research Center for Multimedia Software School of Computer ScienceWuhan UniversityWuhan 430072China Hubei Key Laboratory of Multimedia and Network Communication Engineering Wuhan UniversityWuhan 430072China Research Institute of Wuhan University in Shenzhen China

The head-related transfer function(HRTF)involves the cues for human auditory localization,which turns it into an essential item of virtual auditory display *** practice,the interpolation of HRTF is necessary for the virtual auditory display systems to achieve high spatial *** geometric-based interpolation methods are generally restrained by the spatial distribution of reference on *** the spatial distribution is sparse,the accuracy of interpolation decreases ***,an interpolation method using the common-pole/zero model and the fitting neural network is ***,we propose a common-pole/zero model to represent HRTFs across multiple subjects,in which the low-dimensional features of the measured HRTFs are ***,for a new spatial direction,we predict the corresponding low-dimensional HRTF with a fitting neural ***,we reconstruct the high-dimensional HRTF from the predicted low-dimensional *** simulation results suggest that the proposed method outperforms other interpolation methods such as Linear_AMBC,Bilinear_AMBC,and the Combination method.

关键词： HRTF interpolation fitting neural network common-pole/zero model

来源：评论

学校读者我要写书评

暂无评论

Interpretability is a Kind of Safety: An Interpreter-based Ensemble for Adversary Defense

arXiv

引用

arXiv 2023年

作者： Wang, Jingyuan Wu, Yufan Li, Mingxuan Lin, Xin Wu, Junjie Li, Chao MOE Engineering Research Center of ACAT School of Computer Science Engineering Beihang University China Beijing Key Laboratory of ESSTCO School of Economics and Management Beihang University Beijing China State Key Laboratory of Software Development Environment Beihang University Beijing100191 China Beijing Advanced Innovation Center for BDBC Beihang University Beijing China

While having achieved great success in rich real-life applications, deep neural network (DNN) models have long been criticized for their vulnerability to adversarial attacks. Tremendous research efforts have been dedicated to mitigating the threats of adversarial attacks, but the essential trait of adversarial examples is not yet clear, and most existing methods are yet vulnerable to hybrid attacks and suffer from counterattacks. In light of this, in this paper, we first reveal a gradient-based correlation between sensitivity analysis-based DNN interpreters and the generation process of adversarial examples, which indicates the Achilles's heel of adversarial attacks and sheds light on linking together the two long-standing challenges of DNN: fragility and unexplainability. We then propose an interpreter-based ensemble framework called X-Ensemble for robust adversary defense. X-Ensemble adopts a novel detection-rectification process and features in building multiple sub-detectors and a rectifier upon various types of interpretation information toward target classifiers. Moreover, X-Ensemble employs the Random Forests (RF) model to combine sub-detectors into an ensemble detector for adversarial hybrid attacks defense. The non-differentiable property of RF further makes it a precious choice against the counterattack of adversaries. Extensive experiments under various types of state-of-the-art attacks and diverse attack scenarios demonstrate the advantages of X-Ensemble to competitive baseline methods. Copyright © 2023, The Authors. All rights reserved.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

科学研究的第五范式——以智能驱动的材料设计为例

引用

engineering 2023年第5期24卷 126-137,I0003,I0004页

作者： Can Leng Zhuo Tang Yi-Ge Zhou Zean Tian Wei-Qing Huang Jie Liu Keqin Li Kenli Li Science and Technology on Parallel and Distributed Processing Laboratory National University of Defense TechnologyChangsha 410073China Laboratory of Software Engineering for Complex Systems National University of Defense TechnologyChangsha 410073China National Supercomputing Center in Changsha Changsha 410082China College of Computer Science and Electronic Engineering Hunan UniversityChangsha 410082China Institute of Chemical Biology and Nanomedicine State Key Laboratory of Chemo/Biosensing and ChemometricsCollege of Chemistry and Chemical EngineeringHunan UniversityChangsha 410082China Department of Applied Physics School of Physics and ElectronicsHunan UniversityChangsha 410082China Department of Computer Science State University of New YorkNew PaltzNY 12561USA

科学正在进入一个新时代——第五范式——它被认为是知识整合到不同领域的主要特征,是基于无所不在的机器学习系统的计算社区中智能驱动的工作。在此,我们通过在天河一号超级计算机系统上构建的催化材料专门设计的典型平台案例,生动地... 详细信息

科学正在进入一个新时代——第五范式——它被认为是知识整合到不同领域的主要特征,是基于无所不在的机器学习系统的计算社区中智能驱动的工作。在此,我们通过在天河一号超级计算机系统上构建的催化材料专门设计的典型平台案例,生动地阐明了第五范式的本质,旨在促进第五范式在其他领域的培养。第五范式平台主要包括模型自动构建(原始数据提取)、指纹自动构建(神经网络特征选择)以及跨学科知识串联的重复迭代(“火山图”)。与分解一起进行的是对迭代中实现的体系结构的性能评估。通过讨论,第五范式的智能驱动平台可以极大地简化和改进研究中极其繁琐和具有挑战性的工作,并通过补偿机器学习中缺少样本和替代一些由于计算资源不足而导致的数值计算来实现数值计算与机器学习之间的相互反馈,从而加速探索过程。在数据驱动的学科中,跨学科专家的协同作用和对动态数据需求的急剧增长仍然是一个挑战。我们相信,对第五范式平台的一瞥可以为其在其他领域的应用铺平道路。

关键词：机器学习自动构建天河一号数据驱动神经网络动态数据知识整合科学研究

来源：评论

学校读者我要写书评

暂无评论

DeepSolo++: Let Transformer Decoder with Explicit Points Solo for Multilingual Text Spotting

arXiv

引用

arXiv 2023年

作者： Ye, Maoyuan Zhang, Jing Zhao, Shanshan Liu, Juhua Liu, Tongliang Du, Bo Tao, Dacheng School of Computer Science National Engineering Research Center for Multimedia Software Institute of Artificial Intelligence Hubei Key Laboratory of Multimedia and Network Communication Engineering Wuhan University Wuhan China School of Computer Science Faculty of Engineering The University of Sydney Australia JD Explore Academy *** China College of Computing & Data Science Nanyang Technological University Singapore

End-to-end text spotting aims to integrate scene text detection and recognition into a unified framework. Dealing with the relationship between the two sub-tasks plays a pivotal role in designing effective spotters. Although Transformer-based methods eliminate the heuristic post-processing, they still suffer from the synergy issue between the sub-tasks and low training efficiency. Besides, they overlook the exploring on multilingual text spotting which requires an extra script identification task. In this paper, we present DeepSolo++, a simple DETR-like baseline that lets a single Decoder with Explicit Points Solo for text detection, recognition, and script identification simultaneously. Technically, for each text instance, we represent the character sequence as ordered points and model them with learnable explicit point queries. After passing a single decoder, the point queries have encoded requisite text semantics and locations, thus can be further decoded to the center line, boundary, script, and confidence of text via very simple prediction heads in parallel. Furthermore, we show the surprisingly good extensibility of our method, in terms of character class, language type, and task. On the one hand, our method not only performs well in English scenes but also masters the transcription with complex font structure and a thousand-level character classes, such as Chinese. On the other hand, our DeepSolo++ achieves better performance on the additionally introduced script identification task with a simpler training pipeline compared with previous methods. Extensive experiments on public benchmarks demonstrate that our simple approach achieves better training efficiency compared with Transformer-based models and outperforms the previous state-of-the-art. For example, on ICDAR 2019 ReCTS for Chinese text, our method boosts the 1-NED metric to a new record of 78.3%. On ICDAR 2019 MLT, DeepSolo++ achieves absolute 5.5% H-mean and 8.0% AP improvements on joint detection an

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Cramér-Rao Bound Minimization for a Passive Sensing Multi-ISAC System

引用

IEEE Transactions on Vehicular Technology 2025年

作者： Chen, Jian Chen, Qikai Deng, Yansha Jia, Jie Yin, Baoxin Wang, Xingwei Northeastern University School of Computer Science and Engineering Shenyang110819 China Northeastern University Engineering Research Center of Security Technology of Complex Network System Ministry of Education Shenyang110819 China Northeastern University Key Laboratory of Intelligent Computing in Medical Image Ministry of Education Shenyang110819 China King's College London LondonWC2R 2LS United Kingdom

The integrated sensing and communication (ISAC) system can simultaneously provide communication and radar sensing, effectively improving spectrum utilization. However, existing research mainly focuses on the mono-static sensing ISAC system, which has shortcomings in limited sensing range and accuracy. We thus investigate a passive sensing multi-ISAC system with multiple communication users and sensing targets, and consider two cases for multi-target sensing without and with prior target knowledge at the base stations (BSs), in which the BSs are interested in estimating the targets' response matrix and channel coefficients, respectively. We also formulate a joint beamforming problem to minimize the Cramér-Rao bound (CRB) of the worst estimation constrained by communication requirements, and obtain the optimal solution with semidefinite relaxation (SDR) assisted transformation. To verify the advantages of our proposed scheme, we consider the bi-static BSs system (B-B) and beamforming with sensing interference cancellation (W-S) as two benchmark schemes. Numerical results illustrate that the proposed scheme obtains better sensing performance and effectively alleviates the effect caused by sensing interference. © 1967-2012 IEEE.

关键词： Cramer-Rao bounds

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：