检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Yang, Ziyun Choy, Kevin Farsiu, Sina The Biomedical Engineering Department Duke University DurhamNC27705 United States The Biomedical Engineering Electrical and Computer Engineering Ophthalmology and Computer Science Departments Duke University DurhamNC27705 United States

Generic object detection is a category-independent task that relies on accurate modeling of objectness. We show that for accurate semantic analysis, the network needs to learn all object-level predictions that appear at any stage of learning, including the pre-defined ground truth (GT) objects and the ambiguous decoy objects that the network misidentifies as foreground. Yet, most relevant models focused mainly on improving the learning of the GT objects. A few methods that consider decoy objects utilize loss functions that only focus on the single-response, i.e., the loss response of a single ambiguous pixel, and thus do not benefit from the wealth of information that an object-level ambiguity learning design can provide. Inspired by the human visual system, which first discerns the boundaries of ambiguous regions before delving into the semantic meaning, we propose a novel loss function, Spatial Coherence Loss (SCLoss), that incorporates the mutual response between adjacent pixels into the widely-used single-response loss functions. We demonstrate that the proposed SCLoss can gradually learn the ambiguous regions by detecting and emphasizing their boundaries in a self-adaptive manner. Through comprehensive experiments, we demonstrate that replacing popular loss functions with SCLoss can improve the performance of current state-of-the-art (SOTA) salient or camouflaged object detection (SOD or COD) models. We also demonstrate that combining SCLoss with other loss functions can further improve performance and result in SOTA outcomes for different applications. The codes will be released to https://***/TBD upon acceptance. Copyright © 2024, The Authors. All rights reserved.

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

Smart Watch to Avoid Face Touching During Pandemic Condition 3

Smart Watch to Avoid Face Touching During Pandemic Condition

引用

3rd International Conference on Innovative Practices in Technology and Management, ICIPTM 2023

作者： Vimala, S. Raman, Ramakrishnan Gowrishankkar, V. Bharathiraja, N. Thamarai, I. Kumar, Ashok Panimalar Engineering College Department of Electrical and Electronics Chennai India Symbiosis Institute of Business Management Pune & Symbiosis International Pune India School of Management Sri Krishna College of Technology Chennai India VelTech MultiTech Department of Biomedical Engineering Chennai India Panimalar Engineering College Department of Computer Science Chennai India BanasthaliVidyapith Department of Computer Science Rajasthan India

ISBN: (纸本)9798350336238

The most dangerous Coronavirus, COVID-19, is the source of this pandemic illness. This illness was initially identified in Wuhan, China, in December 2019, and currently sweeping the globe. The virus spreads quickly because it is so simple to transmit from one person to another. Fever is one of the obvious signs of COVID-19 and is one of its prevalent symptoms. The mucosal areas, such as the nose, eyes, and mouth, are among the most significant ways to catch this virus. In order to prevent and track the corona virus infection, this research suggests a face-touching detection and self-health report monitoring system. Their hygiene will immediately improve thanks to this system. In this pandemic circumstance, people use their hands in dirty environments like buses, trains, and other surfaces, where the virus can remain active for a very long time. With an accelerometer and a pulse oximeter sensor, this system alerts the user when they are carrying their hands close to their faces. © 2023 IEEE.

关键词： Viruses

来源：评论

学校读者我要写书评

暂无评论

An Environmentally Friendly Fishing Android-Based Mobile Device App for Penghu Islands (Pescadores), Taiwan

An Environmentally Friendly Fishing Android-Based Mobile Dev...

引用

Consumer Technology (ICCT-Pacific), International Conference on

作者： Hung-Tu Lin Ke-Qing Yan Ying-Zun Fang Xiang-Rui Huang Liang-Bi Chen Department of Computer Science and Information Engineering National Penghu University of Science and Technology Penghu Taiwan Department of Electrical Engineering Master Program in Electrical Engineering and Computer Science National Penghu University of Science and Technology Penghu Taiwan

ISBN: (数字)9798331504120

ISBN: (纸本)9798331504137

Penghu Islands (Pescadores), Taiwan's outlying islands surrounded by the sea, boast rich fishing resources that attract many anglers. However, successful fishing requires more than just gear: it involves understanding the varying fish species available in different locations, times, and climates, as well as adhering to relevant laws and regulations. To address these needs, this paper proposes an environmentally friendly fishing Android-based mobile device App specifically for the Penghu Islands. The proposed Android-based mobile device App aims to gather essential information, including local fishing regulations, aquatic safety tips, and important precautions. Its goal is to equip fishermen with the resources necessary to fish successfully and with peace of mind.

关键词： Sea measurements Length measurement Fish Regulation Mobile applications Tides Protection Smart phones Meteorology Lenses

来源：评论

学校读者我要写书评

暂无评论

A comprehensive evaluation of large language models in mining gene relations and pathway knowledge

引用

Quantitative Biology 2024年第4期12卷 360-374页

作者： Muhammad Azam Yibo Chen Micheal Olaolu Arowolo Haowang Liu Mihail Popescu Dong Xu Department of Electrical Engineering and Computer Science University of MissouriColumbiaMissouriUSA Bond Life Sciences Center University of MissouriColumbiaMissouriUSA Institute for Data Science and Informatics University of MissouriColumbiaMissouriUSA Department of Biomedical Informatics Biostatistics and Medical EpidemiologyUniversity of MissouriColumbiaMissouriUSA

Understanding complex biological pathways,including gene–gene interactions and gene regulatory networks,is critical for exploring disease mechanisms and drug *** literature curation of biological pathways cannot keep up with the exponential growth of new discoveries in the ***-scale language models(LLMs)trained on extensive text corpora contain rich biological information,and they can be mined as a biological knowledge *** study assesses 21 LLMs,including both application programming interface(API)-based models and open-source models in their capacities of retrieving biological *** evaluation focuses on predicting gene regulatory relations(activation,inhibition,and phosphorylation)and the Kyoto Encyclopedia of Genes and Genomes(KEGG)pathway *** indicated a significant disparity in model ***-based models GPT-4 and Claude-Pro showed superior performance,with an F1 score of 0.4448 and 0.4386 for the gene regulatory relation prediction,and a Jaccard similarity index of 0.2778 and 0.2657 for the KEGG pathway prediction,***-source models lagged behind their API-based counterparts,whereas Falcon-180b and llama2-7b had the highest F1 scores of 0.2787 and 0.1923 in gene regulatory relations,*** KEGG pathway recognition had a Jaccard similarity index of 0.2237 for Falcon-180b and 0.2207 for *** study suggests that LLMs are informative in gene network analysis and pathway mapping,but their effectiveness varies,necessitating careful model *** work also provides a case study and insight into using LLMs das knowledge *** code is publicly available at the website of GitHub(Muh-aza).

关键词： biomedical text mining gene-gene interaction KEGG pathway large language model

来源：评论

学校读者我要写书评

暂无评论

Using a 3D-Printed Waveguide Filter with Ridge Resonators as a Dielectric Permittivity Sensor

Using a 3D-Printed Waveguide Filter with Ridge Resonators as...

引用

Microwave, MTT-S International Symposium

作者： Francesco Romano Nicolò Delmonte Luca Perregrini Maurizio Bozzi Department of Electrical Computer and Biomedical Engineering University of Pavia Pavia Italy

This paper presents the use of a waveguide filter with ridge resonators that exhibits transmission zeros to implement a dielectric permittivity sensor. Thanks to its geometry, the electric field is intense under the ridge resonators, where the dielectric slab to be characterized is located. The operation principle is based on the evaluation of the frequency shift of transmission zeros, and exhibits high sensitivity and tolerance to vertical positioning errors. The designed sensor operates around the frequency of 10 GHz. The manufacturing process is based on the 3D printing with plastic material and the subsequent metallization. A set of commercial dielectric laminates have been adopted to validate the performance of the sensor.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A user interface to communicate interpretable AI decisions to radiologists

A user interface to communicate interpretable AI decisions t...

引用

Medical Imaging 2023: Image Perception, Observer Performance, and Technology Assessment

作者： Ou, Yanchen Jessie Barnett, Alina Jade Mitra, Anika Schwartz, Fides Regina Chen, Chaofan Grimm, Lars Lo, Joseph Y. Rudin, Cynthia Department of Computer Science Duke University Durham United States Department of Radiology Duke University Durham United States School of Computing and Information Science University of Maine Orono United States Department of Biomedical Engineering Duke University Durham United States Department of Electrical and Computer Engineering Duke University Durham United States

ISBN: (数字)9781510660403

ISBN: (纸本)9781510660397

Tools for computer-aided diagnosis based on deep learning have become increasingly important in the medical field. Such tools can be useful, but require effective communication of their decision-making process in order to safely and meaningfully guide clinical decisions. Inherently interpretable models provide an explanation for each decision that matches their internal decision-making process. We present a user interface that incorporates the Interpretable AI Algorithm for Breast Lesions (IAIA-BL) model, which interpretably predicts both mass margin and malignancy for breast lesions. The user interface displays the most relevant aspects of the model's explanation including the predicted margin value, the AI confidence in the prediction, and the two most highly activated prototypes for each case. In addition, this user interface includes full-field and cropped images of the region of interest, as well as a questionnaire suitable for a reader study. Our preliminary results indicate that the model increases the readers' confidence and accuracy in their decisions on margin and malignancy. © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.

关键词： Risk assessment

来源：评论

学校读者我要写书评

暂无评论

SoC-Oriented Implementation of Machine Learning Based Breast Cancer Classification Algorithm 12

SoC-Oriented Implementation of Machine Learning Based Breast...

引用

12th Mediterranean Conference on Embedded Computing, MECO 2023

作者： Saeed, Abdelrahman Tawfik, Ayman Mostafa, Hassan Khalil, Ahmed Hussein Cairo University Electronics and Communications Engineering Department Giza12613 Egypt Ajman University Department of Electrical and Computer Engineering United Arab Emirates University of Science and Technology Nanotechnology and Nanoelectronics Program Zewail City of Science and Technology Giza Egypt

ISBN: (纸本)9798350322910

Convolutional Neural Networks (CNN) have drawn the attention of researchers in the medical imaging field. Many researchers have exploited CNN for breast cancer detection. This study provides an Internet of Things (IoT) friendly implementation of CNN for breast cancer detection. To achieve faster time to Market, Deep-learning Processing Unit (DPU) on Field programmable Gate Array (FPGA) is adopted for the CNN hardware implementation. CNN inference on the proposed system achieves a 1.6x speed-up factor and 91.5% reduction in energy consumption compared to the conventional general-purpose multi-core Central Processing Unit (CPU). © 2023 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Deep Semantic Segmentation of Echocardiographic Images using Vision Transformers

Deep Semantic Segmentation of Echocardiographic Images using...

引用

IEEE Symposium (IUS) Ultrasonics

作者： Edoardo Bosco Filippo Casula Marco Cotogni Claudio Cusano Giulia Matrone Department of Electrical Computer and Biomedical Engineering University of Pavia Pavia Italy

Deep learning (DL) methods have revolutionized image segmentation by providing tools to automatically identify structures within images, with high levels of accuracy. In particular, Convolutional Neural Networks (CNN), such as the U-Net and its variants, have achieved remarkable results in many segmentation tasks. Only recently, Vision Transformer (ViT)-based models have emerged and in some cases demonstrated to outperform CNNs in semantic segmentation tasks. However, transformers typically require larger amounts of data for training as compared to CNNs. This can result in a significant drawback given the time-consuming nature of collecting and annotating data, especially in a clinical setting. In particular, only a few studies involved the application of ViT networks for ultrasound image *** this study, we propose one of the earliest applications of ViT-based architectures for segmenting the left heart chambers in 2D echocardiographic images. Indeed, the identification of cardiac structures, like e.g. heart chambers, can be used to derive relevant quantitative parameters, such as atrial and ventricular volumes, the ejection fraction, *** trained and tested several ViT models using the publicly available CAMUS dataset, composed by cardiac image sequences from 500 patients, with the corresponding mask labels of the left ventricle and left atrium. ViT networks performances were then compared to an implementation of the U-Net. We demonstrate how, on this type of data, recent ViT variants can reach and even outperform CNNs, despite the limited data availability.

关键词：

来源：评论

学校读者我要写书评

暂无评论

1st-Order Error-Feedback Sampling-Rate Reconfigurable Noise-Shaping SAR ADC for Multi-Channel CMOS Front-End ASICs for Space Applications

1st-Order Error-Feedback Sampling-Rate Reconfigurable Noise-...

引用

Annual IEEE Northeast Workshop on Circuits and Systems (NEWCAS)

作者： Rashid Karim Marco Grassi Piero Malcovati Department of Electrical Computer and Biomedical Engineering University of Pavia Pavia Italy

This paper proposes a $1^{\text{st}}$-order, error feedback (EF), sampling-rate $(fs)$ reconfigurable Noise-Shaping Successive-Approximation-Register (NS-SAR) Analog-to-Digital-Converter (ADC) for multi-channel CMOS front-end ASICs for X-rays detectors in space applications. The core of the design is an 8-bits Asynchronous-SAR (ASAR) ADC, equipped with a unique variable rising-edge only delay-cell, which generates all required clock signals internally using a single external low duty-cycle sampling signal. The EF NS function has been achieved by introducing a $1^{\text{st}}$-order Finite-Impulse-Response (FIR) loop filter with a duty-cycled OP-AMP. Designed in a $0.35\mu\mathrm{m}$ CMOS process at $3.3\mathrm{~V}$ supply voltage, the proposed NS-SAR ADC achieves an ENOB of ${\bf 1 1. 8}$ bits and a peak SNDR of $72\mathrm{~dB}$ at a typical $fs$ of 40 kHz with an Oversampling-Ratio (OSR) of 8 in schematic-level simulations, consuming an average power of $37\mu\mathrm{W}$. The sampling frequency of the NS-SAR ADC is reconfigurable together with the power consumption while maintaining the same performance. For a fair comparison of the proposed high input signal swing ($2\mathrm{~V}$) ADC at 3.3 V supply with low input signal swing ADCs at lower supply, normalized versions (FoM WN , FoM SN ) of FoM W and $FoM_{S}$ have been proposed. The $ADC$ achieves its best $FoM_{WN}$ of $115.34\mathrm{fJ}/\mathrm{C}$-step and $FoM_{SN}$ of $186.31\mathrm{~dB}$ at its $\operatorname{maximum}f_{S}(178\mathrm{kHz})$ at an OSR of 8.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Smart Optofluidic Sensing Platform Ensuring Patients' Safety During Parenteral Nutrition Administration

A Smart Optofluidic Sensing Platform Ensuring Patients' Safe...

引用

Conference on Lasers and Electro-Optics Europe (CLEO EUROPE)

作者： Valentina Bello Elisabetta Bodo Sabina Merlo Department of Electrical Computer and Biomedical Engineering University of Pavia Pavia Italy

ISBN: (纸本)9798350345995

Parenteral nutrition (PN) is a lifesaving treatment for a large number of patients suffering from different pathologies, from cancer to intestinal failure, from eating disorders to inflammatory bowel disease. In PN, liquid nutritive drugs are injected into the patient's body intravenously through an infusion delivery pump. As reported in the scientific literature, among all medical treatments, PN is the most commonly prone one to human errors [1]. In particular, very often wrong PN mixtures (PNMs) are administered. The consequences can be very dangerous to the patient's health, leading to death in the most severe cases. Despite this fact, currently no standard safety protocols or control devices are implemented to prevent medication errors in PN. Hence, in the framework of the DSF (Digital Smart Fluidics - project No. 1175234, founded by POR FESR 2014–2020), we have developed an optofluidic sensing platform to distinguish different types of transparent commercial PNMs on the basis of their refractive index (RI). Indeed, each mixture contains different concentrations of glucose, amino acids, and electrolytes, determining different values of their RI. The instrumental configuration of the sensor (Fig 1(a)) features a laser diode that generates a red beam (wavelength of 670 nm) travelling obliquely through a plastic cuvette channel containing the PNM under test. Light is then back-reflected by a mirror applied to the back of the cuvette, and finally exits the channel at a position that depends on the RI of the PNM. The position of the output light spot is easily detected with a linear position sensitive detector (PSD) [2]. The two photo-currents $I_{1}$ and $I_{2}$ generated at the extremities of the sensitive area are given as the PSD outputs: by proper normalization, it is possible to retrieve the measured light beam position as $p_{PSD}=L/2\times(I_{1}-I_{2})/(I_{1}+I_{2})$ , where $L$ is the PSD length. The sensor response is then retrieved, after defining t

关键词： Detectors Diode lasers Light beams Mirrors Refractive index Sensors

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：