检索结果-内蒙古大学图书馆

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

作者： Shuai Yu Jun Liu Yi Yu Wei Li School of Computer Science and Technology Donghua University Shanghai China National Institute of Informatics (NII) Tokyo Japan School of Computer Science and Technology Fudan University Shanghai China Shanghai Key Laboratory of Intelligent Information Processing Fudan University Shanghai China

Extracting the melody of a singing voice is an essential task within the realm of music information retrieval (MIR). Recently, transformer based models have drawn great attention in the field of MIR. However, due to the expensive computation cost and extensive parameters, it is difficult to train and deploy a transformer-based model for practical singing melody extraction. In this paper, we propose a simple yet effective scalable sparse transformer for singing melody extraction. To be specific, we first propose to employ a sparse transformer to reduce computation cost and the amount of parameters. Then, we proposed to scale the self-attention region of the sparse transformer in the spectrogram to obtain more accurate performance. Moreover, we propose to combine a scalable sparse transformer (S 2 Former) with CNN-based model to extract global and local features in the spectrogram. The proposed scalable transformer model can achieve a better balance between a standard transformer and a sparse transformer. To better fuse the features from transformer and CNN, we further propose a transformer-CNN fusion (TCF) module to combine significant features from transformer and CNN. The proposed model obtains state-of-the-art results on several public datasets. The conducted experiments confirm the effectiveness of the model we proposed.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Exploring Data Geometry for Continual Learning

Exploring Data Geometry for Continual Learning

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Zhi Gao Chen Xu Feng Li Yunde Jia Mehrtash Harandi Yuwei Wu Beijing Key Laboratory of Intelligent Information Technology School of Computer Science & Technology Beijing Institute of Technology China Guangdong Laboratory of Machine Perception and Intelligent Computing Shenzhen MSU-BIT University China Department of Electrical and Computer Systems Eng. Monash University and Data61 Australia

Continual learning aims to efficiently learn from a non-stationary stream of data while avoiding forgetting the knowledge of old data. In many practical applications, data complies with non-Euclidean geometry. As such, the commonly used Euclidean space cannot gracefully capture non-Euclidean geometric structures of data, leading to in-ferior results. In this paper, we study continual learning from a novel perspective by exploring data geometry for the non-stationary stream of data. Our method dynamically expands the geometry of the underlying space to match growing geometric structures induced by new data, and pre-vents forgetting by keeping geometric structures of old data into account. In doing so, making use of the mixed cur-vature space, we propose an incremental search scheme, through which the growing geometric structures are en-coded. Then, we introduce an angular-regularization loss and a neighbor-robustness loss to train the model, capa-ble of penalizing the change of global geometric structures and local geometric structures. Experiments show that our method achieves better performance than baseline methods designed in Euclidean space.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Two-Stage OD Flow Prediction for Emergency in Urban Rail Transit

引用

IEEE Transactions on intelligent Transportation Systems 2024年第1期25卷 920-928页

作者： Zhu, Guangyu Ding, Jiacun Wei, Yun Yi, Yang Xu, Sendren Sheng-Dong Wu, Edmond Q. Beijing Jiaotong University Key Lab. of Transport Industry of Big Data Application Technologies for Comprehensive Transport The Beijing Research Center of Urban Traffic Information Sensing and Service Technologies Beijing100044 China Beijing Mass Transit Railway Operation Corporation Ltd. Beijing100014 China Yangzhou University College of Information Engineering Yangzhou225127 China National Taiwan University of Science and Technology Automation and Control Center The Graduate Institute of Automation and Control Taipei106335 Taiwan Shanghai Jiao Tong University Key Laboratory of System Control and Information Processing Ministry of Education of China The Shanghai Engineering Research Center of Intelligent Control and Management Department of Automation Shanghai200240 China

Urban rail transit (URT) is vulnerable to natural disasters and social emergencies including fire, storm and epidemic (such as COVID-19), and real-time origin-destination (OD) flow prediction provides URT operators with important information to ensure the safety of URT system. However, hindered by the high dimensionality of OD flow and the lack of supportive information reflecting the real-time passenger flow changes, study in this area is at the beginning stage. A novel model consisting of two stages is proposed for OD flow prediction. The first stage predicts the inflows of all stations by Long Short-Term Memory (LSTM) in real time, where the dimension is reduced compared with predicting OD flows directly. In the second stage, the notion of separation rate, namely, the proportion of inbound passengers bounding for another station, is estimated. Finally, The OD flow is predicted by multiplying the inflow and separation rate. Experiments based on Hangzhou Metro dataset show the proposed model outperforms the contrast model in weighted mean average error (WMAE) and weighted mean square error (WMSE). Results also suggest that the proposed prediction model performs better on weekdays than on weekends, and with greater accuracy on larger OD flows. © 2000-2011 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Balance Multi-Head Attention based on Software and Hardware Co-design 9

Balance Multi-Head Attention based on Software and Hardware ...

引用

9th IEEE International Conference on Cyber Security and Cloud computing and 8th IEEE International Conference on Edge computing and Scalable Cloud, CSCloud-EdgeCom 2022

作者： Xu, Dian Hu, Wei Liu, Fang Fan, Zimeng Shi, Qingsong Wuhan University of Science and Technology Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System College of Computer Science Wuhan China Wuhan University Department of Information Engineering Wuhan Institute of City School of Computer Science Wuhan China Zhejiang University College of Computer Science Hangzhou China

ISBN: (数字)9781665480666

ISBN: (纸本)9781665480666

Recently, the Transformer-based models have achieved leading results in many research areas such as natural language processing and computer vision. However, since Transformer-based models have a huge computational complexities, optimization for Transformer is the focus of current research. the core of Transformer models is the attention module, and in this paper we propose a software and hardware co-designed attention module, which reduces the computational units of the multi-head attention module by balancing the The inference time of the attention module is reduced by balancing the computational units of multiple attention modules. At the same time, we design the corresponding structure on FPGA, which accelerate 14.6x inference time in comparison with the original attention module implementation on GPU. We applied this attention module to multiple visual Transformer models and tested the accuracy difference between the balanced multi-head attention module-based model and the original model on CIFAR-10, and obtained a result of 80.46%, which lower from the original version by 0.18%. © 2022 IEEE.

关键词： Field programmable gate arrays (FPGA)

来源：评论

学校读者我要写书评

暂无评论

BADA-LAT: Efficient Local Attention Transformer for Chinese Named Entity Recognition with Boundary and LLM-Based Data Augmentation 27th

BADA-LAT: Efficient Local Attention Transformer for Chinese ...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Qiu, Xiaoping Yang, Ke Du, Shiling School of Computing and Artificial Intelligence Southwest Jiaotong University No. 999 Xi’an Road Sichuan Chengdu611756 China Tangshan Institute of Southwest Jiaotong University No. 6 Jingqu Road Hebei Tangshan063000 China National United Engineering Laboratory of Integrated and Intelligent Transportation No. 111 North Section 1 Second Ring Road Sichuan Chengdu610031 China Sichuan Key Laboratory of Manufacturing Industry Chain Collaboration and Information Support Technology No. 999 Xi’an Road Sichuan Chengdu610016 China Engineering Research Center of Sustainable Urban Intelligent Transportation No. 999 Xi’an Road Sichuan Chengdu611756 China Sichuan Institute of Industrial Software Technology No. 61 Sha’wan Road Sichuan Chengdu610031 China

ISBN: (纸本)9783031784941

Progress in Chinese Named Entity Recognition (CNER) has highlighted lexicon-based methods that use word information to boost performance. However, these methods neglect two crucial aspects: the regularity of word boundary and the characteristics of the NER task. To address these shortcomings, we introduce BADA-LAT, a Transformer architecture that incorporates word boundary information and introduce local attention computation. It can enhance entity boundary recognition and concentrate the attention weight of the target token on adjacent characters and matched words. Furthermore, to mitigate the issue of class imbalance, we augment the original training data using large language model (LLM). Our method outperforms other lexicon-based ones, as shown in experiments on four Chinese datasets. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： boundary augmentation Chinese NER data augmentation local attention

来源：评论

学校读者我要写书评

暂无评论

Multiscale Spatial Sparse Unmixing for Remotely Sensed Hyperspectral Imagery

Multiscale Spatial Sparse Unmixing for Remotely Sensed Hyper...

引用

IEEE International Symposium on Geoscience and Remote Sensing (IGARSS)

作者： Jiajun Zheng Huqing Liang Shaoquan Zhang Fan Li Pengfei Lai Shengqian Wang Chengzhi Deng Jiangxi Province Key Laboratory of Water Information Cooperative Sensing and Intelligent Processing Nanchang Institute of Technology Nanchang China Guangzhou Municipal Construction Group Co. LTD Guangzhou China

Spectral unmixing is a crucial aspect of hyperspectral image processing. Given the low spatial resolution of hyperspectral remote sensing sensors, combined with the complexity and diversity of actual ground objects, hyperspectral remote sensing images often contain numerous mixed pixels, which make spectral unmixing a challenging task. Recent advancements in spectral libraries have shown promising results for decomposing mixed pixels in hyperspectral remote sensing images. Sparse unmixing, a semi-supervised unmixing strategy, avoids the drawbacks of blind source unmixing algorithms, which may extract virtual endmembers with no physical meaning. In this paper, we propose the multiscale spatial sparse unmixing (MSSU) algorithm, which utilizes the signal adaptive spatial multiscale unmixing of the over-segmentation method to decompose the complex unmixing problem. Furthermore, weighting factors are introduced to extract spatial information from the spectral image. The experimental results obtained from simulated hyperspectral datasets reveal the great potential of the proposed algorithm in unmixing.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Collaborative Consistency Autoencoder Hyperspectral Unmixing Using Deep Image Prior

Collaborative Consistency Autoencoder Hyperspectral Unmixing...

引用

IEEE International Symposium on Geoscience and Remote Sensing (IGARSS)

作者： Min Huang Mengxiong Tang Fan Li Shaoquan Zhang Shengqian Wang Ningyuan Zhang Chengzhi Deng Jiangxi Province Key Laboratory of Water Information Cooperative Sensing and Intelligent Processing Nanchang Institute of Technology Nanchang China Guangzhou Municipal Construction Group Co. LTD Guangzhou China

In the field of hyperspectral unmixing, deep learning has received increasing attention due to its powerful learning and data representation capabilities. Autoencoder is a popular technique for unmixing. Recently, an autoencoder-based depth image prior algorithm has been proposed for hyperspectral unmixing, which employs geometric methods to extract endmembers. The performance of this network solely focuses on estimating the abundance of images, and it has achieved good unmixing results. However, the depth image only employs one core network, which makes it highly vulnerable to noise and can lead to unstable unmixing outcomes. To address the aforementioned issue, this paper proposes a collaborative consistency autoencoder-based hyperspectral unmixing approach with deep image prior (CCAUDIP). For the proposed CCAUDIP model, it adopts two autoencoders to cooperatively handle the same input data to enhance the generalization ability of the network. Additionally, a consistency constraint is introduced to restrict the abundance outputs of the two core autoencoders and improve the robustness of the network. The experimental results show that the CCAUDIP method can achieve better unmixing results compared to other advanced unmixing algorithms.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Mongolian Medicine Named Entity Recognition via Dictionary-Based Synonym Generalization

Mongolian Medicine Named Entity Recognition via Dictionary-B...

引用

IEEE International Conference on Cloud computing and Intelligence Systems (CCIS)

作者： Si Qin Feilong Bao Uuganbaatar Dulamragchaa College of Computer Science Inner Mongolia University Huhhot China National & Local Joint Engineering Research Center of Intelligent Information Processing Technology for Mongolian Huhhot China Inner Mongolia Key Laboratory of Mongolian Information Processing Technology Huhhot China Institute of Mathematics and Digital Technology Mongolian Academy of Sciences Ulaanbaatar Mongolia

The task of named entity recognition in Mongolian medicine poses challenges due to the diversity and complexity of terminology and herbal names. Variations in names used across different literature sources and regions hinder accurate recognition, as traditional dictionary-based methods often fail to encompass all synonyms and variants. To address this issue, this paper presents a lexicon-based synonym generalization method. By leveraging lexical resources in Mongolian, this approach extends the synonyms of common herbs and terms, incorporating additional synonyms and related words into the named entity recognition system. Consequently, the system becomes capable of recognizing a wider range of names and variants. In this study, the authors construct the Mongolian Medicine Named Entity Recognition Datasets (MMNER), comprising five entity types and 44,968 entities. The datasets are developed using the proposed synonym generalization framework. Subsequently, a series of deep learning models are employed to conduct experimental comparisons on the MMNER datasets. The experimental results demonstrate that by enhancing synonym coverage, the recognition of Mongolian medicine entities in text achieves a significantly improved accuracy, with an F1 value of 93.66%. These findings signify a noteworthy enhancement in the performance of Mongolian medicine named entity recognition.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Cross-Modal Object Tracking via Modality-Aware Fusion Network and A Large-Scale Dataset

arXiv

引用

arXiv 2023年

作者： Liu, Lei Zhang, Mengya Li, Cheng Li, Chenglong Tang, Jin The Information Materials and Intelligent Sensing Laboratory of Anhui Province Anhui Provincial Key Laboratory of Multimodal Cognitive Computation School of Artificial Intelligence Anhui University Hefei230601 China The Information Materials and Intelligent Sensing Laboratory of Anhui Province Key Laboratory of Intelligent Computing and Signal Processing The Ministry of Education Anhui Provincial Key Laboratory of Multimodal Cognitive Computation School of Computer Science and Technology Anhui University Hefei230601 China The School of Computer Science and Technology Anhui University Hefei230601 China

Visual tracking often faces challenges such as invalid targets and decreased performance in low-light conditions when relying solely on RGB image sequences. While incorporating additional modalities like depth and infrared data has proven effective, existing multi-modal imaging platforms are complex and lack real-world applicability. In contrast, near-infrared (NIR) imaging, commonly used in surveillance cameras, can switch between RGB and NIR based on light intensity. However, tracking objects across these heterogeneous modalities poses significant challenges, particularly due to the absence of modality switch signals during tracking. To address these challenges, we propose an adaptive cross-modal object tracking algorithm called Modality-Aware Fusion Network (MAFNet). MAFNet efficiently integrates information from both RGB and NIR modalities using an adaptive weighting mechanism, effectively bridging the appearance gap and enabling a modality-aware target representation. It consists of two key components: an adaptive weighting module and a modality-specific representation module. The adaptive weighting module predicts fusion weights to dynamically adjust the contribution of each modality, while the modality-specific representation module captures discriminative features specific to RGB and NIR modalities. MAFNet offers great flexibility as it can effortlessly integrate into diverse tracking frameworks. With its simplicity, effectiveness, and efficiency, MAFNet outperforms state-of-the-art methods in cross-modal object tracking. To validate the effectiveness of our algorithm and overcome the scarcity of data in this field, we introduce CMOTB, a comprehensive and extensive benchmark dataset for cross-modal object tracking. CMOTB consists of 61 categories and 1000 video sequences, comprising a total of over 799K frames. We believe that our proposed method and dataset offer a strong foundation for advancing cross-modal object tracking research. The dataset, toolkit, a

关键词： Large dataset

来源：评论

学校读者我要写书评

暂无评论

Automation 5.0: The Key to Systems Intelligence and Industry 5.0

引用

IEEE/CAA Journal of Automatica Sinica 2024年第8期11卷 1723-1727页

作者： Ljubo Vlacic Hailong Huang Mariagrazia Dotoli Yutong Wang Petros A.Ioannou Lili Fan Xingxia Wang Raffaele Carli Chen Lv Lingxi Li Xiaoxiang Na Qing-Long Han Fei-Yue Wang Institute of Intelligent and Integrated Systems and the School of Engineering and Built Environment Griffith UniversityNathanQLD 4111Australia Department of Aeronautical and Aviation Engineering The Hong Kong Polytechnic UniversityHong KongChina Department of Electrical and Information Engineering Polytechnic of Bari70126 BariItaly State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of AutomationChinese Academy of SciencesBeijing 100190and also with the Qingdao Academy of Intelligent IndustriesQingdao 266114China Department of Electrical Engineering-Systems University of Southern CaliforniaLos AngelesCA 90007 USA School of Automation Beijing Institute of TechnologyBeijing 100081China State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of AutomationChinese Academy of SciencesBeijing 100190 School of Artificial Intelligence University of Chinese Academy of SciencesBeijing 100049 Beijing Huairou Academy of Parallel Sensing Beijing 101499China School of Mechanical and Aerospace Engineering Nanyang Technological UniversitySingapore 639798Singapore Department of Electrical and Computer Engineering Purdue School of Engineering and TechnologyIndiana University-Purdue University IndianapolisIndianapolisIN 46202 USA Department of Engineering University of CambridgeCB21TN CambridgeU.K. School of Science Computing and Engineering TechnologiesSwinburne University of TechnologyMelbourne VIC 3122Australia State Key Laboratory for Management and Control of Complex Systems Chinese Academy of SciencesBeijing 100190 School of Artificial Intelligence University of Chinese Academy of SciencesBeijing 100049China Dazhou Artificial Intelligence Institute Dazhouand the Faculty of Innovation EngineeringMacao University of Science and TechnologyMacao 999078China

AUTOMATION has come a long way since the early days of mechanization,i.e.,the process of working exclusively by hand or using animals to work with *** rise of steam engines and water wheels represented the first generation of industry,which is now called Industry Citation:***,***,***,***,***,***,***,***,***,***,***,Q.-***,and F.-***,“Automation 5.0:The key to systems intelligence and Industry 5.0,”IEEE/CAA ***,vol.11,no.8,pp.1723-1727,Aug.2024.

关键词： Automation machinery intelligence

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：