检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Wang, Xiaoqin Chen, Chen Lan, Rushi Liu, Licheng Liu, Zhenbing Zhou, Huiyu Luo, Xiaonan Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin University of Electronic Technology China College of Electrical and Information Engineering Hunan University China School of Computing and Mathematical Sciences University of Leicester United Kingdom

Unsupervised hashing has attracted much attention for binary representation learning due to the requirement of economical storage and efficiency of binary codes. It aims to encode high-dimensional features in the Hamming space with similarity preservation between instances. However, most existing methods learn hash functions in manifold-based approaches. Those methods capture the local geometric structures (i.e., pairwise relationships) of data, and lack satisfactory performance in dealing with real-world scenarios that produce similar features (e.g. color and shape) with different semantic information. To address this challenge, in this work, we propose an effective unsupervised method, namely Jointly Personalized Sparse Hashing (JPSH), for binary representation learning. To be specific, firstly, we propose a novel personalized hashing module, i.e., Personalized Sparse Hashing (PSH). Different personalized subspaces are constructed to reflect category-specific attributes for different clusters, adaptively mapping instances within the same cluster to the same Hamming space. In addition, we deploy sparse constraints for different personalized subspaces to select important features. We also collect the strengths of the other clusters to build the PSH module with avoiding over-fitting. Then, to simultaneously preserve semantic and pairwise similarities in our proposed JPSH, we incorporate the proposed PSH and manifold-based hash learning into the seamless formulation. As such, JPSH not only distinguishes the instances from different clusters, but also preserves local neighborhood structures within the cluster. Finally, an alternating optimization algorithm is adopted to iteratively capture analytical solutions of the JPSH model. We apply the proposed representation learning algorithm JPSH to the similarity search task. Extensive experiments on four benchmark datasets verify that the proposed JPSH outperforms several state-of-the-art unsupervised hashing algorithms. Cop

关键词： Hash functions

来源：评论

学校读者我要写书评

暂无评论

Performance Analysis of Direction of Arrival Estimation Based on Deep Learning 2020

Performance Analysis of Direction of Arrival Estimation Base...

引用

4th International Conference on Digital Signal processing, ICDSP 2020

作者： Chen, Min Mao, Xingpeng Gong, Yi School of Electronics and Information Engineering Harbin Institute of Technology Shenzhen Engineering Laboratory of Intelligent Information Processing for IoT Southern University of Science and Technology Harbin China School of Electronics and Information Engineering Harbin Institute of Technology Key Laboratory of Marine Environment Monitoring and Information Processing Ministry of Industry and Information Technology Harbin China Shenzhen Engineering Laboratory of Intelligent Information Processing for IoT Southern University of Science and Technology Shenzhen China

ISBN: (纸本)9781450376877

In this paper, a new efficient direction of arrival (DOA) estimation approach based on the deep neural networks (DNN) is proposed, in which a nonlinear mapping that relates the outputs of the receiving antennas with its associated DOA is learning by using the DNN-based network. The novel network architecture is divided into two stages, the detection phase and the DOA estimation phase. Additional detection network attached in our structure dramatically reduces the size of the training set. It has been shown that the proposed method not only can achieve reasonably high DOA estimation accuracy, but also can reduce the computational complexity required by traditional superresolution DOA estimation algorithms such as multiple signal classification (MUSIC). The computer simulation results are performed to investigate the generalization and effectiveness of the proposed approach in different scenarios. © 2020 ACM.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Traditional Mongolian-to-Cyrillic Mongolian Conversion Method Based on the Combination of Rules and Transformer

Traditional Mongolian-to-Cyrillic Mongolian Conversion Metho...

引用

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

作者： Muhan Na Feilong Bao Weihua Wang Guanglai Gao Uuganbaatar Dulamragchaa College of Computer Science Inner Mongolia University Hohhot China Inner Mongolia Key Laboratory of Mongolian Information Processing Technology Inner Mongolia University Hohhot China National & Local Joint Engineering Research Center of Intelligent Information Processing Technology for Mongolian Inner Mongolia University Hohhot China Mongolian Academy of Sciences Institute of Mathematics and Computer Science Ulaanbaatar Mongolia

Mongolian words are composed of stems and suffixes, resulting in a large vocabulary of Mongolian and many Out-of-vocabulary (OOV) words. Using only a dictionary cannot convert Traditional Mongolian OOV words. Therefore, this paper proposes a Traditional Mongolian to Cyrillic Mongolian conversion method based on combining rules and Transformer. This method uses rules based approach for converting in-vocabulary words and the Transformer model to convert OOV words. Experimental results show that using a Transformer outperforms the previous best OOV words conversion model, and the conversion method based on the combination of rules and Transformer has WER of 13.18%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Self-decoupling 5G MIMO Antenna via Grounding for Mobile Phones

Self-decoupling 5G MIMO Antenna via Grounding for Mobile Pho...

引用

IEEE Asia-Pacific Conference on Antennas and Propagation (APCAP)

作者： Wei Zhou Junwei Qi Yingsong Li College of Information and Communication Engineering Harbin Engineering University (HEU) China College of Computer Science and Technology HEU China Key Laboratory of Intelligent Computing Signal Processing Ministry of Education Anhui University Hefei China

This paper presents a self-decoupling high isolation terminal antenna for mobile phones which can applied to 5G multiple-input-multiple-output (MIMO) communication systems. The significant decoupling effect is achieved by introducing only one ground point in each asymmetric antenna pair with an interval of very close (2mm). Through the simulations and optimizations, it can be obtained that the bandwidth of the two antennas is different owing-to the asymmetric structures of the constructed MIMO antenna. The -10 dB resonant frequency ranges of the two antennas are 4.65 - 5.17 GHz and 4.45 - 5.58 GHz, respectively, but the coincidence bandwidth of the two antennas covers the 5G band of 4.8 - 5 GHz, and the coupling of the proposed MIMO antenna in the same bandwidth is lower than -18dB as well.

关键词： Couplings Wireless communication 5G mobile communication Grounding Resonant frequency Bandwidth Mobile antennas

来源：评论

学校读者我要写书评

暂无评论

Point, Segment and Count: A Generalized Framework for Object Counting

Point, Segment and Count: A Generalized Framework for Object...

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Zhizhong Huang Mingliang Dai Yi Zhang Junping Zhang Hongming Shan Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University Shanghai China School of Cyber Science and Engineering Sichuan University Chengdu China Frontiers Center for Brain Science Key Laboratory of Computational Neuroscience and Brain-Inspired Intelligence Institute of Science and Technology for Brain-inspired Intelligence & MOE Fudan University Shanghai China

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

Class-agnostic object counting aims to count all objects in an image with respect to example boxes or class names, a.k.a few-shot and zero-shot counting. In this paper, we propose a generalized framework for both few-shot and zero-shot object counting based on detection. Our framework combines the superior advantages of two foundation models without compromising their zero-shot capability: (i) SAM to segment all possible objects as mask proposals, and (ii) CLIP to classify proposals to obtain accurate object counts. However, this strategy meets the obstacles of efficiency over-head and the small crowded objects that cannot be localized and distinguished. To address these issues, our framework, termed PseCo, follows three steps: point, segment, and count. Specifically, we first propose a class-agnostic object localization to provide accurate but least point prompts for SAM, which consequently not only reduces computation costs but also avoids missing small objects. Furthermore, we propose a generalized object classification that leverages CLIP image/text embeddings as the classifier, following a hierarhical knowledge distillation to obtain discriminative classifications among hierarchical mask proposals. Extensive experimental results on FSC-147, COCO, and LVIS demonstrate that PseCo achieves state-of-the-art performance in both few-shot/zero-shot object counting/detection.

关键词： Location awareness Image segmentation Computer vision Accuracy Costs Power capacitors Pattern recognition

来源：评论

学校读者我要写书评

暂无评论

Octopus-Inspired Multichannel Tactile Sensor for Enhanced Underwater Material Identification

SSRN

引用

SSRN 2024年

作者： Hao, Yutao Sun, Yanshuo Wen, Jing Gao, Xiaobo Wang, Yutong Zhu, Zhiyuan Wang, Zhong Lin Chen, Baodong Chongqing Key Laboratory of Nonlinear Circuits and Intelligent Information Processing College of Electronic and Information Engineering Southwest University Chongqing400715 China Beijing Institute of Nanoenergy and Nanosystems Chinese Academy of Sciences Beijing101400 China School of Nanoscience and Engineering University of Chinese Academy of Sciences Beijing100049 China Research Institute of Technology of Shougang Group Beijing100043 China The Key Laboratory of High Temperature Electromagnetic Materials and Structure of MOE Wuhan University of Science and Technology Wuhan430081 China Georgia Institute of Technology AtlantaGA30332-0245 United States

The sustainable development of the ocean requires sensors capable of detecting objects in underwater or high humidity conditions. However, traditional sensors struggle in complex underwater environments due to signal attenuation, biofouling, and flow interference, which seriously affect the performance and reliability. Inspired by the tactile system of octopus suckers, we developed a tactile sensor that mimics the structure of octopus tentacle suckers, ingeniously harnessing triboelectric tactile receptors (TTRs) to emulate the mechanism of cephalopod-specific chemoreceptors (CRs), aiming to address the challenging problem of underwater object recognition. Additionally, the superhydrophobic treatment increases the microstructure of the sensor surface, effectively mitigating environmental interference and improved underwater performance, resulting in a 67% increase in voltage output, a sensitivity of 0.195 V kPa-1, and a remarkable response time of only 85 ms. Most importantly, we have constructed an underwater material identification system (UMIS) to achieve 98% accuracy by integrating machine learning, which enables precise identification and quantitative sensing of underwater objects, offering novel insights and directions for the intelligence and autonomy of underwater robots. © 2024, The Authors. All rights reserved.

关键词： Shellfish

来源：评论

学校读者我要写书评

暂无评论

Short-term Portfolio Optimization using Doubly Regularized Exponential Growth Rate

Short-term Portfolio Optimization using Doubly Regularized E...

引用

International Conference on Computer Supported Cooperative Work in Design

作者： Quanlong Guan Jinneng He Zhao-Rong Lai Yuyu Zhou Quming Jiang Ziliang Chen College of Information Science and Technology Jinan University Guangzhou China Guangdong Institute of Smart Education Jinan University Guangzhou China Guangdong Key Laboratory of Data Security and Privacy Preserving Guangzhou China Guangdong-Macao Advanced Intelligent Computing Joint Laboratory Zhuhai China Chutian Dragon Co. Ltd Dongguan China School of Intelligent Systems Science and Engineering Jinan University China

ISBN: (数字)9798350349184

ISBN: (纸本)9798350349191

In the realm of short-term portfolio optimization, the integration of machine learning with exponential growth rate techniques is gaining prominence. This paper introduces a novel approach for short-term portfolio optimization, termed Short-term Portfolio Optimization using Doubly Regularized EGR (SPODR), to address the challenges posed by limited data availability. SPODR utilizes radial basis functions for the effective identification of market trends, enabling improved stock market forecasts. The approach uniquely combines ℓ 1 and ℓ 2 -regularization, adhering to empirical financial principles, to strike a balance between risk and return in short-term portfolios. A key aspect of SPODR is addressing the complexity of its ElasticNet-like objective, which poses a challenge for traditional methods due to its online learning nature. To overcome this, we have developed an algorithm based on the log barrier interior-point method. This algorithm is adept at efficiently optimizing portfolio allocation, taking into account the specific constraints inherent in our approach. Extensive comparative experiments across five benchmark datasets demonstrate that SPODR significantly outperforms existing short-term portfolio optimization models. It achieves a right balance between return and risk. Furthermore, SPODR showcases efficient computational speed, enhancing its applicability in real-world financial settings.

关键词： Computational modeling Benchmark testing Market research Robustness Regulation Computational efficiency Resource management

来源：评论

学校读者我要写书评

暂无评论

Attribute-Guided Pedestrian Retrieval: Bridging Person Re-ID with Internal Attribute Variability

Attribute-Guided Pedestrian Retrieval: Bridging Person Re-ID...

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Yan Huang Zhang Zhang Qiang Wu Yi Zhong Liang Wang Center for Research on Intelligent Perception and Computing (CRIPAC) State Key Laboratory of Multimodal Artificial Intelligence Systems (MAIS) Chinese Academy of Sciences Institute of Automation Beijing China School of Artificial Intelligence University of Chinese Academy of Sciences (UCAS). School of Electrical and Data Engineering University of Technology Sydney Australia School of Information and Electronics Beijing Institute of Technology China

ISBN: (数字)9798350353006

ISBN: (纸本)9798350353013

In various domains such as surveillance and smart retail, pedestrian retrieval, centering on person re-identification (Re-ID), plays a pivotal role. Existing Re-ID methodologies often overlook subtle internal attribute variations, which are crucial for accurately identifying individuals with changing appearances. In response, our paper introduces the Attribute-Guided Pedestrian Retrieval (AGPR) task, focusing on integrating specified attributes with query images to refine retrieval results. Although there has been progress in attribute-driven image retrieval, there remains a notable gap in effectively blending robust Re-ID models with intra-class attribute variations. To bridge this gap, we present the Attribute-Guided Transformer-based Pedestrian Retrieval (ATPR) framework. ATPR adeptly merges global ID recognition with local attribute learning, ensuring a co-hesive linkage between the two. Furthermore, to effectively handle the complexity of attribute interconnectivity, ATPR organizes attributes into distinct groups and applies both inter-group correlation and intra-group decorrelation regularizations. Our extensive experiments on a newly estab-lished benchmark using the RAP dataset [32] demonstrate the effectiveness of ATPR within the AGPR paradigm.

关键词： Couplings Pedestrians Correlation Surveillance Image retrieval Focusing Benchmark testing

来源：评论

学校读者我要写书评

暂无评论

Static Gesture Recognition and Distance Estimation Using a Capacitive Proximity Sensor Array

SSRN

引用

SSRN 2023年

作者： Ye, Yong Liu, Yuting Du, Chunhui Deng, Jiahao Qiu, Chenguang School of Electronics and Information Engineering The Information Materials and Intelligent Sensing Laboratory of Anhui Province The Key Laboratory of Intelligent Computing & Signal Processing Ministry of Education The Key Laboratory of Electromagnetic Environmental Sensing of Anhui Higher Education Institutes Anhui University No. 111 Jiulong Road Anhui Province Hefei230601 China Institute of high energy physics Chinese academy of sciences The School of Mechatronical Engineering Beijing Institute of Technology Beijing100081 China Tai Yuan institute Co. Ltd. of China coal technology and engineering group Shanxi Province Taiyuan030006 China East China Institute of Photo-Electronic IC Jiangsu Province Suzhou215163 China

In this paper, we present a capacitive proximity sensor (CPS) array for static gesture recognition and distance estimation. A 4×4 electrode matrix is employed as the sensing element that is configured to 14 electrode operation patterns to obtain 14 capacitive sensors to percept objects within their sensing region. These 14 capacitances are fused into one feature vector which is used as the input of machine learning (ML) classifier for gesture recognition and neural network or regression model for distance estimation. In our experiments, we selected 10 static gestures to verify the proposed method. We compared the performance of 8 ML classifiers and their variations with training data size. The results show that the Bagging achieves the highest average accuracy of 93.2% and the KNN shows a good performance even with a small training data size. The results in distance estimation experiment demonstrate that the RBF network performs a high accuracy. © 2023, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

FoodLogoDet-1500: A dataset for large-scale food logo detection via multi-scale feature decoupling network

arXiv

引用

arXiv 2021年

作者： Hou, Qiang Min, Weiqing Wang, Jing Hou, Sujuan Zheng, Yuanjie Jiang, Shuqiang School of Information Science and Engineering Shandong Normal University Shandong China Key Laboratory of Intelligent Information Processing Institute of Computing Technology Chinese Academy of Sciences Beijing China

Food logo detection plays an important role in the multimedia for its wide real-world applications, such as food recommendation of the self-service shop and infringement detection on e-commerce platforms. A large-scale food logo dataset is urgently needed for developing advanced food logo detection algorithms. However, there are no available food logo datasets with food brand information. To support efforts towards food logo detection, we introduce the dataset FoodLogoDet-1500, a newlarge-scale publicly available food logo dataset, which has 1,500 categories, about 100,000 images and about 150,000 manually annotated food logo objects. We describe the collection and annotation process of FoodLogoDet-1500, analyze its scale and diversity, and compare it with other logo datasets. To the best of our knowledge, FoodLogoDet-1500 is the first largest publicly available high-quality dataset for food logo detection. The challenge of food logo detection lies in the large-scale categories and similarities between food logo categories. For that, we propose a novel food logo detection method Multi-scale Feature Decoupling Network (MFDNet), which decouples classification and regression into two branches and focuses on the classification branch to solve the problem of distinguishing multiple food logo categories. Specifically, we introduce the feature offset module, which utilizes the deformation-learning for optimal classification offset and can effectively obtain the most representative features of classification in detection. In addition, we adopt a balanced feature pyramid in MFDNet, which pays attention to global information, balances the multi-scale feature maps, and enhances feature extraction capability. Comprehensive experiments on FoodLogoDet-1500 and other two popular benchmark logo datasets demonstrate the effectiveness of the proposed method. The code and FoodLogoDet-1500 can be found at https://***/hq03/FoodLogoDet-1500-Dataset. © 2021, CC BY.

关键词： Feature extraction

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：