检索结果-内蒙古大学图书馆

Audio-Text Multimodal Speech Recognition via Dual-Tower Architecture for Mandarin Air Traffic control Communications

引用

computers, Materials & Continua 2024年第3期78卷 3215-3245页

作者： Shuting Ge Jin Ren Yihua Shi Yujun Zhang Shunzhi Yang Jinfeng Yang School of Computer Science and Software Engineering University of Science and Technology LiaoningAnshan114051China Institute of Applied Artificial Intelligence of the Guangdong-Hong Kong-Macao Greater Bay Area Shenzhen Polytechnic UniversityShenzhen518055China Shenzhen Institutes of Advanced Technology Chinese Academy of SciencesShenzhen518055China Industrial Training Centre Shenzhen Polytechnic UniversityShenzhen518055China

In air traffic control communications (ATCC), misunderstandings between pilots and controllers could result in fatal aviation accidents. Fortunately, advanced automatic speech recognition technology has emerged as a promising means of preventing miscommunications and enhancing aviation safety. However, most existing speech recognition methods merely incorporate external language models on the decoder side, leading to insufficient semantic alignment between speech and text modalities during the encoding phase. Furthermore, it is challenging to model acoustic context dependencies over long distances due to the longer speech sequences than text, especially for the extended ATCC data. To address these issues, we propose a speech-text multimodal dual-tower architecture for speech recognition. It employs cross-modal interactions to achieve close semantic alignment during the encoding stage and strengthen its capabilities in modeling auditory long-distance context dependencies. In addition, a two-stage training strategy is elaborately devised to derive semantics-aware acoustic representations effectively. The first stage focuses on pre-training the speech-text multimodal encoding module to enhance inter-modal semantic alignment and aural long-distance context dependencies. The second stage fine-tunes the entire network to bridge the input modality variation gap between the training and inference phases and boost generalization performance. Extensive experiments demonstrate the effectiveness of the proposed speech-text multimodal speech recognition method on the ATCC and AISHELL-1 datasets. It reduces the character error rate to 6.54% and 8.73%, respectively, and exhibits substantial performance gains of 28.76% and 23.82% compared with the best baseline model. The case studies indicate that the obtained semantics-aware acoustic representations aid in accurately recognizing terms with similar pronunciations but distinctive semantics. The research provides a novel modeling pa

关键词： Speech-text multimodal automatic speech recognition semantic alignment air traffic control communications dual-tower architecture

来源：评论

学校读者我要写书评

暂无评论

Focus Your Attention: Multiple Instance Learning with Attention Modification for Whole Slide Pathological Image Classification

引用

IEEE Transactions on Circuits and Systems for Video technology 2025年第6期35卷 5791-5804页

作者： Cheng, Hailun Huang, Shenjin Cai, Linghan Xu, Yangfan Wang, Runming Zhang, Yongbing Tsinghua University Institute of Biopharmaceutical and Health Engineering Shenzhen International Graduate School Shenzhen518055 China Harbin Institute of Technology Faculty of Computing Harbin150001 China Harbin Institute of Technology School of Computer Science and Technology Shenzhen518055 China

computer-aided pathology diagnosis based on whole slide images, which is often formulated as a weakly supervised multiple instance learning (MIL) paradigm. Current approaches generally employ attention mechanisms to aggregate instance-level features. However, the weakly supervised signal and the imbalanced instance distribution often lead to inaccurate attention localization, compromising the performance and generalization capability of the MIL framework. To address these problems, this paper presents a novel MIL framework called FAMIL that focuses on inaccurate attention and refines them. FAMIL adopts a dual-branch structure and incorporates two innovative online data augmentation strategies: attention-based Mixup (ABMix) and attention-based Masking (ABMask). ABMix emphasizes the significance of positive instances, generalizing Mixup in the MIL scenarios, while ABMask flexibly identifies challenging positive instances to optimize the feature representation. Moreover, these two methods are plug-and-play and can be easily embedded into attention-based MIL methods. Extensive experiments on three public benchmarks demonstrate the superiority of our FAMIL, outperforming current state-of-the-art methods. The test AUC for the binary tumor classification can be up to 92.61% over CAMELYON16. And the AUC over the cancer subtype classification can be up to 93.81% and 98.41% on TCGA-NSCLC and TCGA-RCC datasets, respectively. © 1991-2012 IEEE.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Hybrid Optimization Algorithm for Handwritten Document Enhancement

引用

computers, Materials & Continua 2024年第3期78卷 3763-3786页

作者： Shu-Chuan Chu Xiaomeng Yang Li Zhang Václav Snášel Jeng-Shyang Pan College of Computer Science and Engineering Shandong University of Science and TechnologyQingdao266590China School of Information and Control Engineering Qingdao University of TechnologyQingdao266520China Faculty of Electrical Engineering and Computer Science VŠB-TU OstravaOstrava70080Czech Republic Department of Information Management Chaoyang University of TechnologyTaichung41349Taiwan

The Gannet Optimization Algorithm (GOA) and the Whale Optimization Algorithm (WOA) demonstrate strong performance;however, there remains room for improvement in convergence and practical applications. This study introduces a hybrid optimization algorithm, named the adaptive inertia weight whale optimization algorithm and gannet optimization algorithm (AIWGOA), which addresses challenges in enhancing handwritten documents. The hybrid strategy integrates the strengths of both algorithms, significantly enhancing their capabilities, whereas the adaptive parameter strategy mitigates the need for manual parameter setting. By amalgamating the hybrid strategy and parameter-adaptive approach, the Gannet Optimization Algorithm was refined to yield the AIWGOA. Through a performance analysis of the CEC2013 benchmark, the AIWGOA demonstrates notable advantages across various metrics. Subsequently, an evaluation index was employed to assess the enhanced handwritten documents and images, affirming the superior practical application of the AIWGOA compared with other algorithms.

关键词： Metaheuristic algorithm gannet optimization algorithm hybrid algorithm handwritten document enhancement

来源：评论

学校读者我要写书评

暂无评论

Multi-Task Driven Multi-Level Dynamical Fusion for Single-Cell Multi-Omics Cell Type Annotation

Multi-Task Driven Multi-Level Dynamical Fusion for Single-Ce...

引用

2024 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2024

作者： Li, Jiawei Chen, Shizhan Han, Zongbo Li, Wei Tang, Jijun Guo, Fei Tianjin University College of Intelligence and Computing Tianjin300072 China Nankai University College of Artificial Intelligence Tianjin518055 China Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences Shenzhen518055 China Shenzhen University of Advanced Technology Faculty of Computer Science and Control Engineering Shenzhen518107 China Central South University School of Computer Science and Engineering Changsha410083 China

ISBN: (纸本)9798350386226

The emergence of single-cell multi-omics sequencing technology has enabled the simultaneous profiling of diverse omics data within individual cells. It offers a more comprehensive perspective on cellular phenotypes and heterogeneity. However, single-cell multi-omics data are inherently high-dimensional and heterogeneous. Due to technical limitations and scarce starting materials, the data are often affected by noise and dropout effects. To address these challenges, we propose a novel multitask driven multi-level dynamical fusion algorithm for single-cell multi-omics cell type annotation, named scMMDyn. Our approach incorporates reconstruction and classification auxiliary tasks to guide the training of trustworthy modules at both the feature and modality levels. It executes dynamical fusion during these stages and finally achieves cross-modality fusion via an attention mechanism. This method effectively mitigates data quality issues through reconstruction tasks and feature-level dynamical fusion while providing interpretability at both feature and modality levels. Experimental results across diverse single-cell multi-omics datasets show that our method surpasses existing approaches in cell type annotation. © 2024 IEEE.

关键词： cell type annotation multi-task mutli-level dynamical fusion single-cell multi-omics trustworthy

来源：评论

学校读者我要写书评

暂无评论

Improving long-tail classification via decoupling and regularisation

引用

CAAI Transactions on Intelligence technology 2025年第1期10卷 62-71页

作者： Shuzheng Gao Chaozheng Wang Cuiyun Gao Wenjian Luo Peiyi Han Qing Liao Guandong Xu School of Computer Science and Technology Harbin Institute of TechnologyShenzhenChina Peng Cheng Laboratory ShenzhenChina Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies Harbin Institute of TechnologyShenzhenChina Faculty of Engineering and Information Technology University of Technology SydneySydneyNew South WalesAustralia

Real-world data always exhibit an imbalanced and long-tailed distribution,which leads to poor performance for neural network-based *** methods mainly tackle this problem by reweighting the loss function or rebalancing the ***,one crucial aspect overlooked by previous research studies is the imbalanced feature space problem caused by the imbalanced angle *** this paper,the authors shed light on the significance of the angle distribution in achieving a balanced feature space,which is essential for improving model performance under long-tailed ***,it is challenging to effectively balance both the classifier norms and angle distribution due to problems such as the low feature *** tackle these challenges,the authors first thoroughly analyse the classifier and feature space by decoupling the classification logits into three key components:classifier norm(*** magnitude of the classifier vector),feature norm(*** magnitude of the feature vector),and cosine similarity between the classifier vector and feature *** this way,the authors analyse the change of each component in the training process and reveal three critical problems that should be solved,that is,the imbalanced angle distribution,the lack of feature discrimination,and the low feature *** from this analysis,the authors propose a novel loss function that incorporates hyperspherical uniformity,additive angular margin,and feature norm *** component of the loss function addresses a specific problem and synergistically contributes to achieving a balanced classifier and feature *** authors conduct extensive experiments on three popular benchmark datasets including CIFAR-10/100-LT,ImageNet-LT,and iNaturalist *** experimental results demonstrate that the authors’loss function outperforms several previous state-of-the-art methods in addressing the challenges posed by imbalanced and longtailed datasets,t

关键词： computer vision image classification long-tailed data machine learning

来源：评论

学校读者我要写书评

暂无评论

COMPARING WSN PERFORMANCE: TRADITIONAL VS. SDN APPROACH

UPB Scientific Bulletin, Series C: Electrical Engineering an...

引用

UPB Scientific Bulletin, Series C: Electrical engineering and computer science 2024年第4期86卷 81-94页

作者： Hamad, Diyar Jamal Yalda, Khirota Gorgees Okumus, Ibrahim Taner Tapus, Nicolae Faculty of Automatic Control and Computer Science University of Science and Technology POLITEHNICA Bucharest Soran Technical college Erbil Polytechnic University Erbil Iraq Computer Engineering Kahramanmaras Sutcu Imam University Kahramanmaras Turkey Dept. of Computer Science University of Science and Technology POLITEHNICA Bucharest Romania

The proliferation of Wireless Sensor Networks (WSN) in various applications has necessitated the exploration of network architectures that can ensure efficient, scalable, and reliable communication. This study presents a comparative performance evaluation of traditional networking paradigms and Software-Defined Networking (SDN)-based architectures within WSN. Utilizing the Floodlight SDN controller and Mininet-WiFi emulator, we conducted a series of experiments across three different network topologies, varying the number of hosts and access points. The performance metrics focused on are average latency, packet loss percentage, and throughput. The results demonstrate a considerable improvement in all measured parameters under the SDN-based environment compared to the traditional setup. Specifically, SDN environments consistently showcased lower latency, minimal packet loss, and higher throughput, with the most notable difference observed in smaller network configurations. The findings suggest that SDN, with its centralized control plane, can significantly enhance the performance of WSNs, particularly in scenarios demanding rapid scalability and dynamic traffic management. This study underscores the potential of SDN architectures in optimizing WSN operations and lays the groundwork for future research in this domain. © 2024, Politechnica university of Bucharest. All rights reserved.

关键词： Floodlighting

来源：评论

学校读者我要写书评

暂无评论

Sub-terahertz transmissive reconfigurable intelligent surface for integrated beam steering and self-OOK-modulation

引用

Light(science & Applications) 2025年第1期14卷 131-141页

作者： Dongfang Shen Feng Lan Luyang Wang Tianyang Song Munan Yang Tianyu Hu Yueting Li Xiaolei Nie Jiayao Yang Shixiong Liang Hongxin Zeng Hui-Fang Zhang Pinaki Mazumder Ziqiang Yang Yaxin Zhang Tie Jun Cui Sichuan THz Communication Technology Engineering Research Center School of Electronic Science and EngineeringUniversity of Electronic Science and Technology of ChinaChengdu611731China Yangtze Delta Region Institute(Huzhou) University of Electronic Science and Technology of ChinaHuzhou313000China Zhangjiang Laboratory Shanghai201204China School of Microelectronics Tianjin UniversityTianjin300072China Shenzhen Institute for Advanced Study University of Electronic Science and Technology of ChinaShenzhen518100China Department of Electrical Engineering and Computer Science University of MichiganAnn ArborMI48109USA Southeast University Nanjing210096China

Boasting superior flexibility in beam manipulation and a simpler framework than traditional phased arrays,terahertz metasurface-based phased arrays show great promise for 5G-A/6G communication *** with the reflective reconfigurable intelligent surface(reflective RIS),the transmissive RIS(TRIS)offers more feasibility for transceiver multiplexing systems to meet the growing demand for high-performance beam tracking in terahertz communication and radar ***,the terahertz TRIS encounters greater challenges in phase shift,beam efficiency,and complex ***,we propose a sub-terahertz TRIS based on the phase shift via Pancharatnam-Berry(PB)metasurface and self-on-off keying(OOK)modulation via Schottky *** electrically reconfigurable unit cell consists of a column-wise phase resonator and a rectangular *** experimental retrieved equivalent lumped-element circuit model is implemented in joint field-circuit simulations and is validated by experiments.A fabricated prototype demonstrates excellent performance of TRIS with the minimum insertion loss of 2.8 dB for operational states,large bandwidth nearly covering the entire W-band for 1-bit phase shift,deep OOK amplitude modulation of 12 dB,and wide scanning range of±60°with low specular *** further implement an integrated platform combining high-speed beam steering and spatial-light modulation,verifying the point-to-point signal transmissions in different directions using the TRIS *** proposed TRIS with high-performance and cost-effective fabrication makes it a promising solution to terahertz minimalist communication systems,radar,and satellite communication systems.

关键词： reconfigurable reflective resonator

来源：评论

学校读者我要写书评

暂无评论

Automated Recognition and Analysis of Escape Response in Caenorhabditis Elegans

IEEE Transactions on Computational Biology and Bioinformatic...

引用

IEEE Transactions on Computational Biology and Bioinformatics 2025年第3期22卷 1201-1210页

作者： Wenyue Zhan Weiyang Chen Yi Pan School of Cyber Science and Engineering Qufu Normal University Qufu China Faculty of Computer Science and Control Engineering Shenzhen University of Advanced Technology Shenzhen China Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Shenzhen China

Significant progress in medical research in recent years has led to the successful treatment of numerous illnesses. Unfortunately, effective treatments for several neurological disorders—particularly those affecting motor neurons—remain elusive, not only significantly impairing patients' quality of life but also placing a substantial financial burden on both individuals and society. To address this gap, scientists are focusing on model organisms to unravel pathogenic mechanisms and develop treatment strategies. Among various model organisms, Caenorhabditis elegans ( C. elegans) has shown special relevance in neurological disorder research. Escape responses, including reversals and omega turns, are controlled by motor neurons. Thus, the escape response in C. elegans provides researchers with a reliable observational metric to evaluate the functional state of motor neurons. We developed a head and tail position localization model to determine the positions of the nematode's head and tail, enabling automated counting of escape responses. We further clarified the association between escape response and motor neuron function by comparing the escape response of normal nematodes to those with motor neuron deficiencies. Consequently, the automated counting of C. elegans escape responses is a useful tool for assessing the health of motor neurons. It may offer strong backing for the exploration of neurodegenerative disease therapeutics.

关键词： Videos Motors Tail Neurons Head Gray-scale Trajectory Phenotypes Computational modeling Turning

来源：评论

学校读者我要写书评

暂无评论

Character recognition competition for street view shop signs

引用

National science Review 2023年第6期10卷 27-31页

作者： Jingqun Tang Weidong Du Bin Wang Wenyang Zhou Shuqi Mei Tao Xue Xing Xu Hai Zhang Shanghai Byte Dance Technology Co. Ltd School of Control Science and Engineering Zhejiang University School of Computer Science and Technology University of Science and Technology of China School of Computer Science and Technology Tsinghua Univeristy Shenzhen Tencent Technology Co. Ltd School of Computer Science and Engineering University of Electronic Science and Technology of China Pazhou Laboratory (Huangpu) School of Mathematics Northwest University

INTRODUCTION Store signboards provide important information in street view images,andcharacter recognition in natural scenes is an important research direction in computer *** view store signboard character recognitio...

关键词：

来源：评论

学校读者我要写书评

暂无评论

Tangentially-Magnetized-Slot-PM-Assisted Magnet Saturation Relieving Design for Doubly Salient Reluctance Machine Position Sensorless Drive Electric Propulsion Applications 6

Tangentially-Magnetized-Slot-PM-Assisted Magnet Saturation R...

引用

6th International Conference on Electrical engineering and control Technologies, CEECT 2024

作者： Wang, Weiyu Zhang, Xiaodong Niu, Shuangxia Zhao, Xing Pan, Jianfei Fu, Weinong Great Bay University School of Engineering Dongguan China The Hong Kong Polytechnic University Department of Electrical and Electronic Engineering Hong Kong Hong Kong University of York Department of Electronic Engineering York United Kingdom Shenzhen University College of Mechatronics and Control Engineering Shenzhen China Shenzhen University of Advanced Technology Faculty of Computer Science and Control Engineering Shenzhen China

ISBN: (纸本)9798331528096

For doubly salient reluctance machines, the magnetic saturation effect easily occurs, thereby the saturated inductance with complex analytical expression brings difficulties to the inductance modeling procedure during position sensorless control. To tackle this issue, a direct solution from electrical machine design is introduced to regulate magnetic flux distribution. The key is to insert tangentially-magnetized-slot-permanent-magnet into stator slots, thus relieving the magnetic saturation in the stator teeth. Consequently, unsaturated inductance characteristics can be acquired through the magnetic saturation relieving effect, thereby the inductance modeling effort can be significantly decreased, and such design is friendly for doubly salient reluctance machine position sensorless drive application. © 2024 IEEE.

关键词： Stators

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：