检索结果-内蒙古大学图书馆

Advances in Computation, Communication and Information Technology (ICAICCIT), International Conference on

作者： Durjay Thalisetty Anvitha Kopparthi Snigdha Kambhampati Kanuri Sathvika Suresh Kumar Natarajan Department of Computer Science and Engineering - Data Science JAIN (Deemed-to-be University) Ramanagara India

ISBN: (数字)9798331541217

ISBN: (纸本)9798331541224

Sign language translation is a vital technology for facilitating communication between people who cannot understand sign language and others who have hearing impairments.. While several models have been developed for sign language translation, many face limitations such as inadequate accuracy, inability to handle complex signs, and lack of real-time translation capabilities. This paper presents a Sign language translation system designed help reduce the gap in communication between the deaf and hearing communities. The primary objective is to develop an efficient and accurate translator that creates animations in sign language from spoken language, making communication in real-time more accessible for those who are hard of hearing. The system combines Long Short-Term Memory networks for recognition of speech and Convolutional Neural Networks for image-based sign generation, achieving an impressive accuracy rate of 98%. Through Natural Language processing (NLP), the recognized text is filtered and matched with predefined sign language phrases. If a match is found, the system plays an animation corresponding to the phrase; otherwise, it displays each letter individually. This dual-mode approach ensures flexibility in translation, catering to both common phrases and unique words. The high accuracy and adaptability of this system demonstrate its potential as a powerful instrument for improving accessibility and inclusion in daily communication. Results from experiments show that the suggested model performs better than current methods in terms of accuracy, robustness, and real-time performance. Through rigorous evaluation on benchmark datasets and user studies, the model achieves significant improvements in sign language translation accuracy, effectively closing the communication gap between the general public and those who are hearing impaired. In conclusion, the proposed sign language translation model represents a promising advancement in the field, offering enhanced

关键词： Sign language Adaptation models Translation Accuracy Text recognition Auditory system Animation real-time systems Natural language processing Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Utilizing AI and silver nanoparticles for the detection and treatment monitoring of canker in pomelo trees

引用

SENSORS AND ACTUATORS A-PHYSICAL 2024年 368卷

作者： Thao, Le Quang Kien, Do Trung Thien, Nguyen Duy Bach, Ngo Chi Van Hiep, Vuong Khanh, Dang Gia VNU Univ Sci Fac Phys Hanoi 100000 Vietnam Vietnam Natl Univ Hanoi 100000 Vietnam VNU HUS High Sch Gifted Students Hanoi 100000 Vietnam

Citrus canker, a disease caused by the Xanthomonas citri bacterium, is a serious threat that can cause significant damage to crop. Chemical treatments used by farmers for disease management can lead to detection errors, product misuse, and environmental harm. To address this, we proposed the CitrusNet model, a novel deep learning-based approach for detecting common diseases in citrus, including citrus canker. The model is optimized to run on embedded devices with limited configurations and is based on the MobileVit model. Squeezeand-Excitation (SE) blocks were integrated into the MobileNetV2 blocks, and modifications were made to the baseline model. We employed Self-Supervised learning with the Barlow Twins (SSL-BT) algorithm to pre-train the model, addressing the limitations of labeled datasets for citrus diseases. After pre-training, various classifiers were used to fine-tune the model. An image processing technique was applied to identify affected areas, enabling infection and recovery rate calculations during treatment with a 2-ppm silver nanoparticle (NPs) solution. Research results show the CitrusNet model outperformed other commonly used models for mobile devices when trainable parameters were set at 2.33 M and FLoating-point Operations Per Second (FLOPs) at 750.02 M. The classifier achieved 92.33% accuracy and a 79.06% F1-score using the Support Vector Machine (SVM). Finally, we developed a real-time disease recovery monitoring program on a low-configuration embedded device, using feature maps extracted from the CitrusNet model to monitor silver nanoparticle treatment effects on citrus canker.

关键词： Monitor treatment citrus canker Silver nanoparticle Recovery disease rate CitrusNet Self -supervised learning

来源：评论

学校读者我要写书评

暂无评论

Transfer learning based heart valve disease classification from Phonocardiogram signal

引用

BIOMEDICAL SIGNAL processing AND CONTROL 2023年第1期85卷

作者： Maity, Arnab Pathak, Akanksha Saha, Goutam Indian Inst Technol Kharagpur Dept Elect & Elect Commun Engn Kharagpur 721302 W Bengal India

Physiological conditions that prevent heart valves from functioning precisely to ensure proper blood circulation are known as heart valve disorder (HVD). Detection of HVD is critical as untreated heart valve disease often develops life-threatening cardiac diseases. Typical HVD detection methods, like echocardiography, MRI, and cardiac CT, are costly, complex, and require robust healthcare infrastructure. Although, by simple non-invasive listening to heart sound irregularities, an expert physician can anticipate the signs of HVD from ancient times. Contemporary development suggests that with machine learning-based algorithms, a graphical representation of heart sound, known as the phonocardiogram (PCG), can effectively predict the anomaly in the valvular activity. In recent studies, deep learning-based strategies showed promising results in the PCG classification task but demand extensive resources and training data. This work investigates the merits of transfer learning (TL) using pre-trained convolution neural networks for the automatic PCG classification when data is scarce. With standard time-frequency representations (i.e., spectrogram, log-Mel spectrogram, and scalogram) as input features, audio and image-based pre-trained lightweight models are fine-tuned to categorize the PCG. The proposed YAMNet-based TL method classifies four types of HVD data collected from public heart sound databases and achieves overall accuracy, sensitivity, and specificity of 99.83%, 99.59%, and 99.90%, respectively. Alongside, it classifies the PhysioNet/CinC Challenge 2016 dataset into binary classes with 92.23% accuracy. The study achieves high classification metrics despite data scarcity. It also investigates the proposed method's computational efficiency and robustness against practical noise contamination for performance evaluation in a possible real-life scenario.

关键词： Heart valve disease Transfer learning YAMNet time-frequency representation Spectrogram PCG

来源：评论

学校读者我要写书评

暂无评论

An IoT-enabled real-time overhead view person detection system based on Cascade-RCNN and transfer learning

An IoT-enabled real-time overhead view person detection syst...

引用

作者： Ahmad, Misbah Ahmed, Imran Jeon, Gwanggil Centre for Excellence in Information Technology IMSciences Peshawar 1-A Sector E-5 Phase VII HayatabadPeshawar Pakistan Department of Embedded Systems Engineering Incheon National University Incheon Korea Republic of

Internet of things (IoT) is transforming technological evolution in several practical applications. These applications range from smart cities, smart healthcare to intelligent video surveillance, where the primary interest is person monitoring and detection. The amalgamation of Artificial Intelligence (AI) and IoT-based techniques maintain a balance between computational cost and efficiency that is essential for next-generation IoT networks. In this context, a real-time IoT-enabled people detection system is introduced. The developed system performs image processing task over the cloud using an internet connection, thus reduces the computational cost by processing high-resolution images over the cloud. For person detection, a pre-trained Cascade RCNN, a deep learning approach is used. It is an object detection architecture, seeks to address discrediting performance with increased Intersection over Union (IoU) thresholds. As the architecture is pre-trained with COCO data set and the person body’s appearance in overhead perspective is significantly different;thus, additional training is performed to enhance the detection results. Taking advantage of transfer learning architecture is trained for overhead person images, and the newly trained feature layer is added to the existing architecture. Experimental outcomes reveal that additional training increases the detection architecture’s performance with an accuracy rate of 0.96. © 2021, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Facial image Pre-processing for Enhanced Ethnicity Classification

Facial Image Pre-processing for Enhanced Ethnicity Classific...

引用

Intelligent Technologies (CONIT), International Conference on

作者： Keshav Goel Hitesh Dhiman Himanshu Yadav Shailender Kumar Department of Computer Science and Engineering Delhi Technological University New Delhi India

Facial recognition and ethnicity detection is an important task in various applications such as security, marketing, and social sciences. However, the accuracy of Face recognition and ethnicity detection algorithms can be impacted by various factors such as lighting conditions, noise and irrelevant features in image background, variation in facial expressions, face size, pose and orientation across images. In this study, we present a novel face image preprocessing model in this work that aims to enhance the performance of deep learning models like ResNet, DenseNet and other state of the art deep learning models on facial recognition tasks, specifically ethnicity detection by addressing these issues. Our novel method for preprocessing facial images incorporates several techniques to enhance the quality of the input images. Firstly, we removed the background using deepLab-v3. Secondly, we employed facial landmarks detection and facial pose and orientation correction using facial alignment based on the angle between the eyes. Thirdly, to facilitate race classification, we performed facial cropping to isolate the facial region and other critical features essential for the task. Finally, we ensured consistent lighting conditions across all images using the Multi-scale retinex algorithm. These techniques together enabled the deep learning model to better analyze and classify facial data. To assess the efficacy of our model, we carried out a series of experiments on the UTK Face dataset and compared the outcomes with those obtained when our preprocessing model was not *** experimental outcomes show that our suggested pre-processing model improves the f1-score of ethnicity detection classifiers by up to 6%. The proposed facial image preprocessing model is efficient, fast, and easy to implement, enabling real-time applications, making it appropriate. Our research highlights the importance of preprocessing techniques in enhancing ethnicity detection performance and p

关键词：

来源：评论

学校读者我要写书评

暂无评论

Adversarial domain adaptation for multi-device retinal OCT segmentation

Adversarial domain adaptation for multi-device retinal OCT s...

引用

Medical Imaging Conference - image processing

作者： He, Yufan Carass, Aaron Liu, Yihao Saidha, Shiv Calabresi, Peter A. Prince, Jerry L. Johns Hopkins Univ Dept Elect & Comp Engn Baltimore MD 21218 USA Johns Hopkins Univ Dept Comp Sci Baltimore MD 21218 USA Johns Hopkins Univ Dept Neurol Sch Med Baltimore MD 21287 USA

ISBN: (纸本)9781510633940

deep networks provide excellent image segmentation results given copious amounts of supervised training data (source data). However, when a trained network is applied to data acquired at a different clinical center or on a different imaging device (target data), a significant drop in performance can occur due to the domain shift between the test data and the network training data. To solve this problem, unsupervised domain adaptation methods retrain the model with labeled source data and unlabeled target data. In real practice, retraining the model is time consuming and the labeled source data may not be available for people deploying the model. In this paper, we propose a straightforward unsupervised domain adaptation method for multi-device retinal OCT image segmentation which does not require labeled source data and does not require retraining of the segmentation model. The segmentation network is trained with labeled Spectralis images and tested on Cirrus images. The core idea is to use a domain adaptor to convert target domain images (Cirrus) to a domain that can be segmented well by the already trained segmentation network. Unlabeled Spectralis and Cirrus images are used to train this domain adaptor. The domain adaptation block is used before the trained network and a discriminator is used to differentiate the segmentation results from Spectralis and Cirrus. The domain adaptation portion of our network is fully unsupervised and does not change the previously trained segmentation network.

关键词： unsupervised domain adaptation OCT deep learning segmentation

来源：评论

学校读者我要写书评

暂无评论

A Robust Technique for real-time Face Verification with a Generative Network

A Robust Technique for Real-Time Face Verification with a Ge...

引用

Conference on real-time image processing and deep learning

作者： Akkaya, Ibrahim Batuhan Karaman, Kaan ASELSAN Res Ctr TR-06370 Ankara Turkey

ISBN: (纸本)9781510635807

real-time face verification from a live stream is still an open question, although it is quite popular in recent years. In order to overcome this problem, several bio-metric techniques are widely used for authentication purposes, both in military and civilian areas. It is merely the task of detecting and comparing a candidate face with the other faces in the database to validate whether they are the same person or not. Typically, A face verification pipeline is composed of four stages: Face detection, alignment, recognition, and matching. The faces in the frames of the live stream are detected via a deep neural network (DNN) in the first part. Then, the detected faces are aligned, and another DNN extracts the face features. The feature vector of each face is used for matching with other vectors in the database to validate the identity. New developments in deep learning lead to achieving human-level performance on the aforementioned tasks. However, the networks used in the stages require high computation power. In order to achieve real-time performance in resource-limited devices, lightweight networks should be preferred. Unfortunately, usage of these kinds of networks decreases the detection and recognition performance dramatically in some frames of a live stream. Therefore, the set of feature vectors for an individual, collected from the live stream, contains outliers that complicate obtaining a robust reference feature vector, which is essential for achieving high confidence in verification tasks. In this work, a conditional generative network is utilized for generating these vectors for the given candidate. We conduct the experiments on a real-life scenario for showing the incrementation of performance that is caused by our proposed generative network.

关键词： Face verification generative networks deep learning

来源：评论

学校读者我要写书评

暂无评论

Enhancing incremental deep learning for FCCU end-point quality prediction

引用

INFORMATION SCIENCES 2020年 530卷 95-107页

作者： Zhang, Xu Zou, Yuanyuan Li, Shaoyuan Shanghai Jiao Tong Univ Dept Automat Shanghai 200240 Peoples R China Minist Educ China Key Lab Syst Control & Informat Proc Shanghai 200240 Peoples R China

In the refining process, Fluid Catalytic Cracking Unit (FCCU) end-point quality prediction plays an important role in real-time quality monitoring, optimization and control. Due to the process uncertainties (such as changes of raw materials, abrasion of mechanic components, catalyst deactivation, changes in the external environment, etc), data distribution of process data in FCCU is time-varying, which leads to accuracy degradation of the quality prediction model. Therefore, to deal with the time varying characteristic of process data and avoid accuracy degradation of the quality prediction model, real-time processing should be considered. In this paper, an enhancing incremental deep learning approach is proposed for the online quality prediction of the absorption-stabilization system in FCCU. First, the offline model is built by Stacked Auto Encoder-deep Neural Network (SAE-DNN). To determine whether the data distribution has changed and model modification is needed, a concept drift detector is proposed for the regression problem by defining an error bound in the Statistical Test of Equal Proportions (STEPD). If the model modification is needed, then the top layer of the offline SAE-DNN model is expanded by Random Vector Functional Link (RVFL) structure, and the parameters in the expansion layer is dynamically assigned by the new coming data with the group lasso regularization and the L2 regularization. The proposed approach is validated by predicting the Saturated Vapor Pressure (SVP) of stabilized gasoline in the FCCU. The experimental results show that the proposed approach can deal with the time-varying characteristic of process data and avoid accuracy degradation under process uncertainties. (C) 2020 Elsevier Inc. All rights reserved.

关键词： Fluid catalytic cracking unit Uncertainties time-varying characteristic real-time processing Incremental deep learning

来源：评论

学校读者我要写书评

暂无评论

Multistage real-time Fire Detection Using Convolutional Neural Networks and Long Short-Term Memory Networks

引用

IEEE ACCESS 2021年 9卷 146667-146679页

作者： Manh Dung Nguyen Hoai Nam Vu Duc Cuong Pham Choi, Bokgil Ro, Soonghwan Posts & Telecommun Inst Technol Dept Elect Engn Hanoi 12110 Vietnam Posts & Telecommun Inst Technol Dept Comp Sci Hanoi 12110 Vietnam IVS Vietnam Co Hanoi 12110 Vietnam Kongju Natl Univ Dept Elect Engn Cheonan 31080 South Korea Kongju Natl Univ Dept Informat & Commun Cheonan 31080 South Korea

Fire is one of the most commonly occurring disasters and is the main cause of catastrophic personal injury and devastating property damage. An early detection system is necessary to prevent fires from spreading out of control. In this paper, we propose a multistage fire detection method using convolutional neural networks (CNNs) and long short-term memory (LSTM) networks. In the first stage, fire candidates are detected by using their salient features, such as their color, flickering frequency, and brightness. In the second stage, a pretrained CNN model is used to extract the 2D features of flames that are the input for the LSTM network. In the last stage, a softmax classifier is utilized to determine whether the flames represent a true fire or a nonfire moving object. The experimental results show that our proposed method can achieve competitive performance compared with other state-of-the-art methods and is suitable for real-world applications.

关键词： image color analysis Feature extraction Machine learning algorithms deep learning Computer architecture Microprocessors Heuristic algorithms Fire detection convolutional neural network imageNet long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Neural Architecture Search for Enhancing Action Video Recognition in Compressed Domains

Neural Architecture Search for Enhancing Action Video Recogn...

引用

International Conference on Systems, Signals and image processing, IWSSIP

作者： Pedro Lamkowski Douglas Rodrigues Leandro A. Passos João P. Papa Jurandy Almeida Department of Computing São Paulo State University Bauru São Paulo Brazil Department of Computing Federal University of São Carlos Sorocaba São Paulo Brazil

ISBN: (数字)9798350391886

ISBN: (纸本)9798350391893

Video classification models have become one of the most widely used topics in the computer vision field, encompassing many tasks such as medical, security, industrial, and other applications. Although deep learning models have achieved great results in the video domain, such models are built to operate in the domain of RGB frame sequences. In such models, a prior step is required for decoding video data since the vast majority relies on compressed formats. Nevertheless, large amounts of computational resources are required for decoding, especially in real-time. Researchers have already tackled the task of building networks that work in the compressed domain with promising results but with architectures still very close to those used for the RGB domain. We propose an approach that employs Neural Architecture Search to explore and find the most effective architectures for the compressed domain. Our approach was tested on UCF101 and HMDB51 datasets, obtaining a computationally less complex architecture than similar methods.

关键词： image coding Computational modeling Architecture Computer architecture Streaming media Turning Vectors

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：