检索结果-内蒙古大学图书馆

2024 IEEE Silchar Subsection Annual Conference, SILCON 2024

作者： Karna, Harsh Bardhan Dipta, Debabrata Dash Datta, Bidipta Chandra Pal, Vipin National Institute of Technology Department of Electronics and Instrumentation Engineering Assam Silchar788010 India Institute of Technology Department of Electrical Engineering National Assam Silchar788010 India

ISBN: (纸本)9798331540821

The objective of this paper is to provide a comprehensive overview of brain tumors using different deep-learning models. This study focuses on four deep-learning models, which are Inception V3, MobileNetV2, ResNet152V2, and VGG19, aiming to enhance the accuracy of tumor detection. The dataset utilized in this study was sourced from Kaggle, and categorized into four distinct types, which are pituitary, glioma, meningioma, and no tumor. The f1 score is accessed and precise as well as accurate model is determined by confusion matrix for finding tumor. The enhanced accuracy exhibited by these models can lead us to detect brain tumors at an early stage so that patients can take medical treatment at the first stage of the disease. This technology has the potentiality to revolutionize the field of medical treatment and assist individuals in taking the necessary steps to mitigate losses resulting from tumors. © 2024 IEEE.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

VAG Signal Classification Using Time Domain Statistical Features and Machine Learning 3

VAG Signal Classification Using Time Domain Statistical Feat...

引用

3rd IEEE International Conference on Mobile Networks and Wireless Communications, ICMNWC 2023

作者： Patankar, Sanika Durge, Gaurav Joshi, Ashutosh Jaid, Atharva Kalambe, Khomesh Dhale, Hrishikesh Vishwakarma Institute of Technology Department of Instrumentation and Control Engineering Maharashtra Pune India

ISBN: (纸本)9798350317022

This research paper presents a novel approach for the classification of vibroarthrographic (VAG) signals to aid in the early diagnosis of knee arthritis, a disease whose prevalence is increasing worldwide. VAG signals are non-invasive and cost-effective, making them a promising tool for screening and detection. In this study, a time-domain statistical feature approach is employed to extract relevant features from VAG signals. Analytical measures like Kurtosis, skewness, Standard Deviation have been used in this work. These features are then fed into machine learning algorithms for the automatic detection of knee arthritis. The classification accuracy of 2 algorithms has been compared in this work, namely SVM and Random Forest. The latter resulted in an overall accuracy of 89% and an AUROC of. 90. The results indicate the potential of this approach to provide an accurate and reliable means for early knee arthritis detection, which would allow the patient to be treated ahead of time. © 2023 IEEE.

关键词： Classification Knee arthritis Machine learning signal processing Time-domain statistical features Vibroarthrography

来源：评论

学校读者我要写书评

暂无评论

Design of a High-Sensitive Chipless RFID Sensor for Surface Crack Characterization on Metals 2

Design of a High-Sensitive Chipless RFID Sensor for Surface ...

引用

2nd IEEE International Conference on Futuristic Technologies, INCOFT 2023

作者： Suresh, Setti Preethi, B. Rinu Chakaravarthi, Geetha National Institute of Technology Applied Electromagnetics Laboratory Department of Instrumentation and Control Engineering Tamil Nadu Tiruchirappalli620015 India

ISBN: (纸本)9798350308846

This paper presents the design of a high-sensitivity chipless RFID sensor tag with a modified depolarizing circular patch for surface crack characterization on metals. The sensor response was realized using a cross-polarization reading, and induced dual resonance frequencies from the incident waves. The circular patch resonates at frequencies of 2.766 GHz and 3.204 GHz with four L-shaped tip-loaded resonators as identification encoders. The surface cracks in vertical, horizontal, and diagonal orientations with varying crack widths and depths in sub-millimeter resolution are simulated. It is observed that the designed patch is sensitive to detect a minimum crack resolution of 0.1 mm for varying crack widths and depths. The average computed gradient of simulated sensor upper frequencies is 27.5 MHz and 58.5 MHz for varying crack depths and widths in vertical orientation. Similarly, varying frequency gradients between dual resonances were observed for other orientations. © 2023 IEEE.

关键词： Chipless RFID sensor Radio Frequency Identification Surface crack Wireless NDT

来源：评论

学校读者我要写书评

暂无评论

Circular Ring Microwave Applicator for Superficial Hyperthermia Treatment

Circular Ring Microwave Applicator for Superficial Hyperther...

引用

2023 IEEE Microwaves, Antennas, and Propagation Conference, MAPCON 2023

作者： Alex, Mable Chakaravarthi, Geetha Applied Electromagnetics Laboratory Department of Instrumentation and Control Engineering National Institute of Technology Tamil Nadu Tiruchirappalli620015 India

ISBN: (纸本)9798350328264

This work presents the design of a circular ring applicator operating at a frequency of 434 MHz for superficial hyperthermia treatment of cancer. Simulation study is carried out on a homogeneous model to optimize the parameters of antenna design. The proposed antenna has a bandwidth of 26 MHz at -10 dB, power coupling > 90%, and a penetration depth of 13.67 mm. The specific absorption rate (SAR) of the proposed antenna on homogenous tissue at 5mm depth is 417 W/kg, which is higher compared to other applicators reported in the literature. The performance of the designed antenna is evaluated using a heterogeneous model that is developed from the patient image data. The assessment on the patient derived model indicates that the proposed applicator can be used for hyperthermia treatment of cancer. © 2023 IEEE.

关键词： Breast cancer Chest wall recurrence Hyperthermia Microwave Specific absorption rate

来源：评论

学校读者我要写书评

暂无评论

An Efficient Digitizer for Self-Heating Compensation in Resistive Sensor

An Efficient Digitizer for Self-Heating Compensation in Resi...

引用

2023 International Conference on Recent Advances in Electrical, Electronics and Digital Healthcare Technologies, REEDCON 2023

作者： Joshi, Ashish Dineshkumar Sohail, Shiraz Singh, Gopal Kumar, Nandigama Praveen Islam, Tarikul National Institute of Technology Department of Chemical Engineering Tamil Nadu Tiruchirappalli India National Institute of Technology Department of Instrumentation and Control Engineering Tamil Nadu Tiruchirappalli India New Delhi India

ISBN: (纸本)9781665493826

Self-heating (S-H) error is found in all resistive sensors (RX) when interfaced with an electronic circuit for the output signal. This work presents an efficient digitizer circuit for interfacing the resistive sensor to compensate the S-H error. The digitizer utilizes a dual-slope integrator and a comparator to generate a time sequence output signal, which can be directly fed to the digital pin of a microcontroller for signal processing. To minimize the self-heating error, the input current through RX is kept small in the micro-ampere range. However, a T-network is connected to the integrating capacitor to boost the incoming current in the feedback path and helps in achieving high accuracy, sensitivity, and resolution. This is a novel feature that has not been reported before. The digitizer is simulated for measuring RX in a wide range of 3.63 kΩ to 488.2 kΩ, with corresponding power dissipation of 0.275 mW to 2.03 μW respectively. The current through RX ranges from 275.2 μA -2.04 μA, and the measurement time varies from 4.78 ms to 389.3 ms. If this range of current is fed to a conventional digitizer for integrating the feedback capacitor, the circuit requires a longer measurement time (103.4 ms to 13.76 s). However, with the current boosting approach employed (ranging from 21.07-49.41 times) in this digitizer, the measurement time is reduced. It is to be noted that the conventional digitizer with a measurement time of 4.78 ms to 389.3 ms would require 5800 μA - 101.19 μA to pass through RX which would dissipate 122.11 mW - 4998.88 μW. Thus, this scheme decreases power dissipation by 444 - 2462.4 times which represents a significant improvement. The proposed digitizer produces accurate results with an average error of ± 0.28 %. © 2023 IEEE.

关键词： Error compensation

来源：评论

学校读者我要写书评

暂无评论

Advanced YOLOv8 with CBAM and BiFPN for Assessing Coral Health in Underwater Environments

Advanced YOLOv8 with CBAM and BiFPN for Assessing Coral Heal...

引用

2024 IEEE Silchar Subsection Annual Conference, SILCON 2024

作者： Das, Doneela Mukherjee, Koena National Institute of Technology Silchar Department of Electronics and Instrumentation Engineering Assam Silchar788010 India

ISBN: (纸本)9798331540821

Coral reefs are crucial for marine ecosystems, providing essential ecological and socioeconomic benefits. However, these ecosystems are increasingly threatened by climate change and pollution, exposing the limitations of traditional monitoring methods in precision, scalability, and real-time responsiveness. This study introduces a novel application of the YOLOv8 algorithm, enhanced with the Convolutional Block Attention Module (CBAM) and the Bidirectional Feature Pyramid Network (BiFPN), to address these challenges and advance real-time coral health assessment. Utilizing a comprehensive dataset of 28,500 images, encompassing diverse coral types and conditions, the enhanced YOLOv8 model significantly outperforms its predecessors, including YOLOv5 and YOLOv7. It achieves a mean Average Precision (mAP) of 88.4% at a 0.5 Intersection over Union (IoU), an F1 score of 0.87001, and a processing time of 11.2 ms, while maintaining a compact size of 7.2 MB. These results illustrate a significant improvement of 6.1% over the original YOLOv8 architecture and demonstrate the model's effectiveness in accurately detecting and classifying coral health in challenging underwater environments. This study not only advances AI-driven underwater object detection but also emphasizes the critical role of large, diverse datasets in enhancing model accuracy. The enhanced YOLOv8 model shows significant potential for integration into autonomous underwater vehicles (AUVs), offering a powerful tool for continuous, real-time marine conservation efforts. © 2024 IEEE.

关键词： Reefs

来源：评论

学校读者我要写书评

暂无评论

An experimental analysis of quad wheel autonomous robot location and path planning using Borahsid algorithm with GPS and ZigBee

引用

International Journal of Vehicle Information and Communication Systems 2022年第3期7卷 290-305页

作者： Borah, Siddhanta Kumar, R. Mukherjee, Subhradip Panwala, Fenil C. Lakshmi, A. Prasanna Department of Electronics and Instrumentation Engineering National Institute of Technology Nagaland Nagaland India

This paper presents a hardware system structure and wirelessnavigation system for both localisation and path navigation of a mobile robot implementing a 32-bit ARM processor (LPC2148 Board) into the design process of a mobile robot integrated with GPS and ZigBee wireless communication device. A novel path-navigation algorithm (Borahsid Algorithm) with less complexity is compared with the existing algorithms adopted for mobile robot realistic work. It utilises GPS localisation as well as ZigBee communication. Existing algorithms have higher accuracy in path following, but are a little slower in processing time. The proposed algorithm has faster processing speeds compared with the existing algorithms because of its lower cost function. MATLAB programming language has been used to simulate the mobile robot localisation and path navigation. The results show the effectiveness of the model and the feasibility of the Borahsid Algorithm (BsA). However, the entire control structure is executed and the experimental results were obtained in a real time system by which experimental results authenticates the performance and steadiness of the implemented control system process. © 2022 Inderscience Enterprises Ltd.. All rights reserved.

关键词： MATLAB

来源：评论

学校读者我要写书评

暂无评论

Thermometry Across Switching Oxide Layer in ReRAM Device 1

引用

21st International Workshop on the Physics of Semiconductor Devices, IWPSD 2021

作者： Das, Om Prakash Pandey, Shivendra Kumar Department of Electronics and Instrumentation Engineering National Institute of Technology Silchar Assam Silchar788010 India

ISBN: (数字)9789819715718

ISBN: (纸本)9789819715701

Resistive random access memory (ReRAM) is an emerging non-volatile memory (NVM) technology candidate, having the ability to address the next-generation data storage demand. Even though localized self-heating factors impact the functioning of ReRAM, thorough investigations are limited due to nanoscale geometry constraints. This study employs thermometry across the oxide layer to monitor the generation and distribution of local temperature in both SET and RESET states. A maximum local temperature (Tmax) of 690 K at the filament center is estimated during the low resistive SET state, with a complete cylindrical filament of diameter 0.5 nm upon an applied voltage (Va) of 0.55 V at the top electrode (TE). A transient heat transfer model achieves the current heating time on a nanosecond scale (~ 1 ns). Similarly, during the RESET state with Va = 0.1 V, a gap of 5 nm is considered at the bottom electrode (BE) interface, where a Tmax of 400 K is obtained at the filament tip near BE. The radial temperature distribution is evaluated from the filament to surrounding oxide, where 52.93% and 23.44% local temperature declination are detected at the filament-oxide junction during SET and RESET states, respectively, where the local temperature further declines to room temperature at device walls. Additionally, the local temperature concerning Va is calculated across the positive and negative voltage sweep from an analytical simulation, where the device variables and model parameters have collaborated with an experimental I-V characteristic from a fabricated ReRAM device. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Heat conduction

来源：评论

学校读者我要写书评

暂无评论

Unfalsified Adaptive control of a Robotic System 2nd

Unfalsified Adaptive Control of a Robotic System

引用

2nd International Conference on Emerging Electronics and Automation, E2A 2022

作者： Rasul, Tabassum Mukherjee, Koena Department of Electronics and Instrumentation Engineering National Institute of Technology Silchar Assam Cachar788010 India

ISBN: (纸本)9789819968541

The paper presents a data-driven control method based on the concept of unfalsified control theory. The controller design presented is an adaptive learning-based technique. In this paper, an effective parameter update method is presented by which the system can quickly adapt to parametric changes and uncertainties while fulfilling the desired performance criteria. The method is applied to a robotic arm with unknown parameters for trajectory tracking. The design is tested through simulation in MATLAB/SIMULINK environment, and the effectiveness can be proved by the results presented in the paper. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Autonomous underwater vehicles

来源：评论

学校读者我要写书评

暂无评论

Feature estimation and parameter tuning using H∞ Filter on FastSLAM framework 6

Feature estimation and parameter tuning using H∞ Filter on ...

引用

6th International Conference on Energy, Power and Environment, ICEPE 2024

作者： Shreya, Varanasi Radhika Mishra, Arnav Mukherjee, Koena National Institute of Technology Silchar Department of Electronics and Instrumentation Engineering Assam Silchar788010 India

ISBN: (纸本)9798350379297

Simultaneous Localization and Mapping (SLAM) plays a key role in enabling mobile robots to autonomously navigate and build a map of their environment. Despite its wide range of applications, localization and mapping has its own disadvantages which include computation, feature representation,sensor noise, data association and robustness in dynamic environments. EKF-SLAM and FastSLAM are two frameworks of SLAM. However, they face problems related incorrect priori knowledge of mean, correlation and covariances of process as well as measurement noise as Kalman Filter acts as a minimum covariance estimator. In this paper an Extended H-infinity based FastSLAM is proposed which takes into consideration worst case minimization error and priori knowledge of noise is not essential. The proposed framework is an integration of Rao-Blackwellization Particle filter for proposal distribution and robot pose estimation along with Extended H-Infinity Filter for feature/landmark estimation. Further, the impact of the tuning parameter θ on accuracy is tested. The proposed algorithm is evaluated on a simulated environment and is compared with the existing frameworks. © 2024 IEEE.

关键词： Extended Kalman filters

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：