检索结果-内蒙古大学图书馆

Open-Space Emergency Guiding with Individual Density Prediction Based on Internet of Things Localization

IEEE Transactions on Emerging Topics in Computational Intelligence 2025年第1期9卷 785-797页

作者： Chen, Lien-Wu Huang, Hao-Wei Chen, Yi-Ju Tsai, Ming-Fong Feng Chia University Department of Information Engineering and Computer Science Taichung407 Taiwan National United University Department of Electronic Engineering Miaoli360 Taiwan

This article proposes an open-space emergency guiding (OSEG) framework that explores deep learning techniques to predict individual densities for evacuation based on Internet of Things localization. The OSEG framework adopts Densely Connected Convolutional Networks that can reserve fine-grained features in earlier convolutional layers and extract them in later convolutional layers to efficiently reduce prediction errors. In addition, OSEG can dynamically guide individuals to exits in the shortest time though balancing the evacuation load among exits as individual densities are non-uniform in indoor spaces. According to our review of relevant research, this is the first framework that integrates deep learning for individual density prediction with load-balancing evacuation for emergency guiding in open indoor spaces, where the distance to each exit, capacities of all exits, concurrent moving of evacuees, and distribution of indoor individuals/groups are considered simultaneously. Simulation results show that our approach outperforms existing methods and can balance the evacuation load of each exit for individuals/groups to minimize total evacuation time, which can reduce 50% of total evacuation time at most compared with existing methods. © 2017 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Optimizing Equalizations of FFE, CTLE, and DFE Jointly Through a Single Pulse Response

IEEE Transactions on Signal and Power Integrity

引用

IEEE Transactions on Signal and Power Integrity 2025年 4卷 88-95页

作者： Chen, Yen-Hao Tseng, Chun-I Lin, Ding-Bing National Taiwan University of Science and Technology Department of Electronic and Computer Engineering Taipei106335 Taiwan Inventec Corporation Taoyuan330417 Taiwan

With the increasing data rate of high-speed digital systems, equalization techniques have been widely applied to counteract intersymbol interference and maximize eye opening in today's high-speed serial links. Channel simulations with optimal equalization settings allow engineers to explore design tradeoffs, reduce the need for costly prototypes, and ensure robust performance before manufacturing. Existing equalization optimization methods require either iterative interaction with commercial channel simulators during the optimization process, developing models with extensive training data before optimization, or separately optimizing linear equalizers in the frequency domain. In this article, the method to jointly optimize feedforward equalization, continuous-time linear equalization, and decision feedback equalization at the transmitter, receiver, or both is proposed. The optimization is achieved through the analysis of a nonequalized pulse response, eliminating the computational cost of channel simulations during optimization. Practical examples using non-return-to-zero (NRZ) and pulse amplitude modulation four-level (PAM4) signaling schemes demonstrate the effectiveness of the proposed method. A comparison with the Bayesian optimization approach, a widely discussed method for equalization optimization, shows that while both methods achieve nearly identical optimal eye openings, the proposed method offers significantly higher optimization efficiency. © 2022 IEEE.

关键词： Intersymbol interference

来源：评论

学校读者我要写书评

暂无评论

An encoder-decoder network for crowd counting based on multi-scale attention mechanism

引用

Multimedia Tools and Applications 2025年第3期84卷 1187-1210页

作者： Chuang, Hao-Hsiang Chen, Yi-Cheng Lin, Chang Hong Department of Electronic and Computer Engineering National Taiwan University of Science and Technology No. 43 Keelung Rd. Da’an Dist. Taipei Taiwan

Crowd counting is a challenging computer vision task, which is widely used in video surveillance and public safety applications. With the increase of camera resolution and the complexity of crowd image, it becomes an important problem to predict crowd density and crowd count accurately. Recent CNN-based density estimation methods have shown effectiveness in densely populated scenes. In this paper, we present a novel approach to crowd counting through the development of an Encoder-Decoder Multi-Scale Attention Network. Our approach leverages the robust U-net architecture as the backbone network, strengthened by the strategic integration of an attention mechanism. We adopt a multi-scale attention method to each different layers in the U-net backbone to make the network extract features which focus on the crowds, instead of the background in the images. The attention mechanism and the skip-connections can adjust the weights of feature maps while maintaining features at different scales. Extensive experiments on Shanghaitech Part_A & B and UCF-QNRF dataset demonstrate that our network can achieve better performances with Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) values outperforming existing methodologies: Shanghaitech Part_A (MAE/RMSE: 60.0/104.9), Part_B (MAE/RMSE: 7.8/13.8), and UCF-QNRF (MAE/RMSE: 98.6/179.7). © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Decoding

来源：评论

学校读者我要写书评

暂无评论

Development of Intelligent Remote Sensing Payloads for CubeSat Applications

Development of Intelligent Remote Sensing Payloads for CubeS...

引用

2024 Small Satellites Systems and Services Symposium, 4S 2024

作者： Juang, Jyh-Ching Lee, Pei-Jun Department of Electrical Engineering National Cheng Kung University Taiwan Department of Electronic and Computer Engineering National Taiwan University of Science and Technology Taiwan

ISBN: (纸本)9781510688940

The paper describes the development of a remote sensing payload for CubeSat applications. The inclusion of artificial intelligence into remote sensing payload has been pursued in the IRIS (Intelligent Remote sensing and Internet Satellite) and Lilium CubeSat programs. The paper discusses the development of a series of intelligent remote sensing payloads for CubeSats to address the limitations on computational resources, power, and communication bandwidth. From 2018 to 2021, under the IRIS program, an optical remote sensing payload that contains a high focal-length telescope, a CMOS detector, and the data operation board was developed. An edge intelligence for landslide detection and cloud segmentation is investigated by using MobileNet V2 network. The design has been implemented and tested by using an airborne platform. The design is further refined in the Lilium program with enhanced resolution and frame rate. Functions including auto-focus, automatic white balance, and gain control are also realized in the payload. To account for low-illumination scenario, a zero DCE (zero reference deep curve estimation) is utilized. Further, a light-weight U-net is implemented in the FPGA to achieve cloud detection and segmentation and a YOLO-SPD is implemented for object detection. This self-developed payload has been prototyped and is under tests. The payload will be integrated into the 6U Lilium-2 CubeSat and is planned to be launched in 2025. © 2025 The Authors.

关键词： Proximity sensors

来源：评论

学校读者我要写书评

暂无评论

Trapping-Set-Assisted Decoding for Short-Length Low-Density Parity-Check Codes

引用

IEEE Internet of Things Journal 2025年第9期12卷 11539-11552页

作者： Zamzami, Asif Ali Yu, Zhu-En Wu, Yuh-Tser Lin, Shih-Chun National Taiwan University of Science and Technology Department of Electronic and Computer Engineering Taipei106 Taiwan Universitas Airlangga Department of Robotics and Artificial Intelligence Engineering Surabaya60115 Indonesia National Taiwan University Department of Electrical Engineering Taipei10617 Taiwan Institute of Communications Engineering National Yang Ming Chiao Tung University Hsinchu300 Taiwan National Taiwan University Department of Electrical Engineering Graduate Institute of Communication Engineering Taipei10617 Taiwan

To guarantee that every device in industrial Internet of Things (IoT) is reliable for fast data exchange, ultrareliable and low-latency communication (URLLC) technology from 5G is crucial. To reduce latency, typically, only up to 1000 code blocklength is allowed. Indeed, a block error rate (BLER) of less than 10-5 and latency of less than 1 ms are achievable in 5G URLLC. The short code blocklength feature is also required in satellite IoT. Though low-density parity-check (LDPC) decoding via belief propagation (BP) shows promise for achieving low BLER with long blocklength, its BLER is high in short LDPC codes due to error-prone patterns, with the trapping set (TS) being one of the most significant patterns. To alleviate the damage of TS, which can be identified offline, we propose two decoding algorithms named dynamic TS-assisted reinforced BP (DTS-RBP) and TS-assisted relaxation for neural-network BP (TSR-NNBP), respectively. Compared with previous works, only proposed TS assisted decoders can meet URLLC requirements under 308 code blocklength at a 3 dB signal-to-noise ratio (SNR). Moreover, under 128 code blocklength for satellite IoT, our DTS-RBP decoder provides the best memory-performance tradeoff with approximately the same number of iterations as BP. © 2014 IEEE.

关键词： 5G mobile communication systems

来源：评论

学校读者我要写书评

暂无评论

PULSE: Progressive Utilization of Log-Structured techniques to Ease SSD Write Amplification in B-epsilon-tree 25

PULSE: Progressive Utilization of Log-Structured Techniques ...

引用

30th Asia and South Pacific Design Automation Conference, ASP-DAC 2025

作者： Peng, Huai-De Chen, Yi-Shen Chen, Tseng-Yi Chang, Yuan-Hao Department of Computer Science and Information Engineering National Central University Taiwan Department of Electronic and Computer Engineering National Taiwan University of Science and Technology Taiwan Institute of Information Science Academia Sinica Taiwan

ISBN: (纸本)9798400706356

During the unprecedented expansion of global data, efficient storage solutions are essential for processing massive datasets stored on modern storage devices. B-epsilon-tree (Bϵ-tree) is one of the most well-known techniques that provides a write-optimized structure for database file systems. With the excellent access performance and high energy efficiency of solid-state drives (SSDs), they are expected to yield promising outcomes for large-scale data computation. However, their integration into storage systems has the challenges of write amplification, which impacts SSD endurance and reliability. This work identifies significant write amplification issues with Bϵ-tree implementations on SSDs due to the complicated management of key-value pairs. To mitigate the impact of write amplification, we propose PULSE, a novel scheme that rethinks Bϵ-tree designs by leveraging log-structured techniques optimized for SSDs. Moreover, PULSE integrates auxiliary indexing and a dual flush selector to minimize write amplification. Experimental results demonstrate that PULSE significantly mitigates write amplification by more than 62.6% on SSDs for the representative benchmarks compared to the Bϵ-tree. © 2025 Copyright is held by the owner/author(s). Publication rights licensed to ACM.

关键词： Flash-based SSDs

来源：评论

学校读者我要写书评

暂无评论

Elevated-temperature dry sliding wear behaviors of dual-cathode HiPIMS TiN and TiZrN films

引用

Emerging Materials Research 2025年第1期14卷 98-106页

作者： Lin, Yu-Tse Liu, Wen-Chung Kuo, Chin-Chiuan Department of Power Mechanical Engineering National Formosa University Huwei Yunlin County Taiwan Department of Mechanical and Computer Aided Engineering National Formosa University Huwei Yunlin County Taiwan Graduate Institute of Aeronautical and Electronic Engineering National Formosa University Huwei Yunlin County Taiwan

Titanium–zirconium nitride films with three Zr/Ti ratios (16/35, 22/29, and 25/26) were deposited using Ti and Zr dual-cathode reactive high-power impulse closed-field magnetron sputtering by adjusting the average power ratio of two cathode materials. The sliding wear resistances of the obtained TiZrN films at different temperatures were compared. The results revealed that the ternary TiZrN film exhibited a higher hardness (approximately 35 GPa) and lower elastic modulus (approximately 360–380 GPa) than those of the binary TiN film. This contributed to the lower friction coefficient and wear resistance at room temperature. However, the addition of Zr became disadvantageous with an increase in wear temperature. Above 4008C, the surface hardness of the TiZrN film decreased dramatically with an increase in the Zr fraction. This is attributed to the increased vulnerability of the film to oxidation. Structural discontinuity induced by the inhomogeneous oxidation (a multilayer structure with voids, embedded oxides, and oxynitrides formed in the surface layer) roughened the film surface and reduced the surface strength, which accelerated the wear of TiZrN films at elevated temperature in an ambient atmosphere. Such ternary nitride coatings should not be applied in oxidative high temperature environments without obstructing the oxidation through designed composition and structure. © 2025 Emerald Publishing Limited: All rights reserved.

关键词： Vacuum deposited coatings

来源：评论

学校读者我要写书评

暂无评论

An Identification System for Blood Pressure Values Powered by Deep Learning

An Identification System for Blood Pressure Values Powered b...

引用

2025 International Workshop on Advanced Imaging technology, IWAIT 2025

作者： Chang, Tsai-Rong Lee, Kuo-Chen Fu, Xiang-Ming Department of Electronic Engineering National Kaohsiung University of Science and Technology Kaohsiung807618 Taiwan Department of Computer Science and Information Engineering Southern Taiwan University Tainan71005 Taiwan

ISBN: (数字)9781510688131

ISBN: (纸本)9781510688124

This study presents an innovative application utilizing image recognition technology to automatically identify values displayed on electronic blood pressure *** study uses deep learning, specifically a Mask-RCNN-based algorithm, to accurately detect and recognize numerical readings on monitor *** incorporating advanced image processing techniques such as mask binarization, Canny edge detection, and Hough transform, the system corrects image distortions caused by varying camera angles, transforming them into standardized rectangular formats for precise *** testing under diverse lighting and angle conditions ensured consistent high-precision *** solution demonstrates significant potential for automating medical data management through image recognition technology. © 2024 SPIE.

关键词： Hough transforms

来源：评论

学校读者我要写书评

暂无评论

Improving free space optics performance using QC-LDPC codes and simplified bit flipping decoding in low visibility conditions

引用

Optical and Quantum electronics 2025年第1期57卷 1-18页

作者： Marlina, Lina Pamukti, Brian Liaw, Shien-Kuei Lee, Pei-Jun Kishikawa, Hiroki Graduate Institute of Electro-Optical Engineering National Taiwan University of Science and Technology Taipei Taiwan Department of Electronic and Computer Engineering National Taiwan University of Science and Technology Taipei Taiwan Institute of Post-LED Photonics Tokushima University Tokushima Japan

Currently, free space optics (FSO) is the most promising technology for achieving high data transfer over short to medium or even long distances. Thus, FSO provides wireless line-of-sight (LOS) connectivity in the unlicensed spectrum and can be a high-bandwidth wireless alternative in contrast to fiber optic cabling. Moreover, the FSO offers rapid assembly times and significant cost savings. However, atmospheric weather (e.g., fog, rain, and snow) can degrade the FSO performance, especially in low visibility conditions. Therefore, this research aims to improve the FSO performance using quasi-cyclic low-density parity-check (QC-LDPC) codes in low visibility scenarios or thick turbulence conditions. In addition, a simplified bit-flipping (SBF) algorithm was proposed to decode the information by the receiver side to minimize complexity. Subsequently, the FSO system is simulated using a 1550 nm laser at different data rates of 1 to 20 Gbps data rate and 1-Watt transmitted power. This paper was analyzed using different empirical models of atmosphere attenuation, like the Kim, Kruse, and Naboulsi models. This simulation also uses modulation techniques such as on-off keying, phase shift keying, and quadrature amplitude modulation. Moreover, the looping iteration number of SBF could improve the system’s performance. We also consider using a low-cost PIN photodetector for practical implementation. Our simulation results show that QC-LDPC codes can achieve an average bit error rate of 10-9 in lower visibility than uncoded information. Our research has significant implications for enhancing fiber optic performance in adverse atmospheric phenomena. © The Author(s) 2025.

关键词： Quadrature amplitude modulation

来源：评论

学校读者我要写书评

暂无评论

An Abnormal Behavior Recognition System for Bathrooms with Using Toilet Based on Thermal Image AI Edge Computing technology 16th

An Abnormal Behavior Recognition System for Bathrooms with U...

引用

16th International Conference on Genetic and Evolutionary Computing, ICGEC 2024

作者： Chang, Wan-Jung Liu, Jia-Hong Chen, Yi-Jia Chen, Liang-Bi Department of Electronic Engineering National Kaohsiung University of Science and Technology Kaohsiung807618 Taiwan Department of Electronic Engineering Southern Taiwan University of Science and Technology Tainan710301 Taiwan Department of Computer Science and Information Engineering National Penghu University of Science and Technology Penghu880011 Taiwan

ISBN: (纸本)9789819615346

With the gradual increase in the proportion of the global elderly population, the number of bathroom accidents also increases yearly. Due to the humid environment in bathrooms, the floors are mostly made of tiles, which are relatively prone to slips and falls. However, the existing pose recognition methods mostly use color images, and color images have concerns about personal privacy and are susceptible to lighting changes and complex spatial interference. Therefore, this paper proposes an abnormal behavior recognition system based on thermal image AI edge technology. Use a dangerous posture when using the toilet to avoid accidents in the bathroom, such as using the toilet being discovered too late, increasing the risk. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Differential privacy

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：