检索结果-内蒙古大学图书馆

Multi-task transfer learning deep convolutional neural network: application to computer-aided diagnosis of breast cancer on mammograms

引用

PHYSICS IN MEDICINE AND BIOLOGY 2017年第23期62卷 8894-8908页

作者： Samala, Ravi K. Chan, Heang-Ping Hadjiiski, Lubomir M. Helvie, Mark A. Cha, Kenny H. Richter, Caleb D. Univ Michigan Dept Radiol Ann Arbor MI 48109 USA

Transfer learning in deep convolutional neural networks (DCNNs) is an important step in its application to medical imaging tasks. We propose a multi-task transfer learning DCNN with the aim of translating the 'knowledge' learned from non-medical images to medical diagnostic tasks through supervised training and increasing the generalization capabilities of DCNNs by simultaneously learning auxiliary tasks. We studied this approach in an important application: classification of malignant and benign breast masses. With Institutional Review Board (IRB) approval, digitized screen-film mammograms (SFMs) and digital mammograms (DMs) were collected from our patient files and additional SFMs were obtained from the Digital Database for Screening Mammography. The data set consisted of 2242 views with 2454 masses (1057 malignant, 1397 benign). In single-task transfer learning, the DCNN was trained and tested on SFMs. In multi-task transfer learning, SFMs and DMs were used to train the DCNN, which was then tested on SFMs. N-fold cross-validation with the training set was used for training and parameter optimization. On the independent test set, the multi-task transfer learning DCNN was found to have significantly (p = 0.007) higher performance compared to the single-task transfer learning DCNN. This study demonstrates that multi-task transfer learning may be an effective approach for training DCNN in medical imaging applications when training samples from a single modality are limited.

关键词： digitized screen-film mammography digital mammography computer-aided diagnosis mass transfer learning multi-task learning deep convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

Robust and fast Plant Pathology Prognostics (P³) tool based on deep convolutional neural network

引用

MULTIMEDIA TOOLS AND APPLICATIONS 2022年第5期81卷 7271-7283页

作者： Sasikaladevi, N. SASTRA Deemed Univ Sch Comp Dept CSE Thanjavur TN India

deep learning is emerging as an automatic and accurate model for image classification. Plant diseases are significant threats to food security. Rapid and accurate identification of plant pathology is difficult due to the lack of infrastructure and techniques. The recent advancements of deep learning in computer vision have paved a new horizon for plant pathology diagnosis. Early detection of plant pathology is a demanding task today. This paper proposes a deep convolutional neural network model for the accurate and rapid identification of plant disease. The deep convolutional neural network is designed based on Hypergraph modeling. The plant village dataset covers 38 different classes of 14 other plants. Experimental results show that the proposed model provides the maximum accuracy of 99.7%. Precision, recall, and F1 scores are computed to validate the model. Micro precision and Micro recall analysis are performed to validate the model at the micro-level. Furthermore, it is proved that the proposed model outperforms all the state-of-the-art deep learning models for plant disease detection based on images.

关键词： Plant disease detection deep convolutional neural network Hypergraph neural network Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Obtaining Super-Resolution Satellites Images Based on Enhancement deep convolutional neural network

引用

INTERNATIONAL JOURNAL OF AERONAUTICAL AND SPACE SCIENCES 2021年第1期22卷 195-202页

作者： Keshk, Hatem Magdy Yin, Xu-Cheng Univ Sci & Technol Beijing Beijing Peoples R China Natl Author Remote Sensing & Space Sci Cairo Egypt

Super-resolution reconstruction refers to the technique of reconstructing a high-resolution image from a single or a series of low-resolution images by digital image processing. This technology can not only increase the high-frequency information of the image, but also eliminate the low-resolution. deep Learning has made breakthroughs in modern digital image processing. Compared to traditional algorithms, deep convolutional neural networks (DCNN) achieve superior performance on a series of challenging image-processing problems such as image classification and target detection. Enhancement deep convolutional neural networks (EDCNN) learn through a large number of training samples, obtain relevant information within the image, and then use the information to achieve specific functions. EDCNN also has an excellent performance with remote sensing data. Performance evaluation was made with bicubic and other deep learning methods, EDCNN outperformed other deep learning algorithms.

关键词： Super-resolution Remote sensing deep learning convolutional neural network deep convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

MFIF-DWT-CNN: Multi-focus image fusion based on discrete wavelet transform with deep convolutional neural network

引用

MULTIMEDIA TOOLS AND APPLICATIONS 2024年第4期83卷 10951-10968页

作者： Avci, Derya Sert, Eser Ozyurt, Fatih Avci, Engin Firat Univ Dept Comp Technol Elazig Turkiye Malatya Turgut Ozal Univ Dept Comp Engn Malatya Turkiye Firat Univ Dept Software Engn Elazig Turkiye

A new fusion method based on Multi-Focus Image Fusion Based on Discrete Wavelet Transform with deep convolutional neural network (MFIF-DWT-CNN) is presented to reduce spatial artifacts and blurring effects in edge details and increase the robustness of multifocal image fusion. The main purpose of the MFIF-DWT-CNN approach is to create a new merged image by collecting the required features from the main image. With the MFIF-DWT-CNN approach, information focused on individual images is combined into a single image, resulting in a clearer image. Within the scope of MFIF-DWT-CNN approach, DWT is applied to the image pairs and the obtained images are then given to the CNN architecture. The MFIF-DWT-CNN approach was developed in this study to reduce spatial artifacts and blurring effects in edge details and to increase the robustness of multifocal image fusion. In order to evaluate our proposed MFIF-DWT-CNN method, QMI, QG, QYi QCB evaluations were made on the public data set. From the experimental results, it is seen that the proposed method gives better results in the relevant metrics than the other methods. This demonstrated the effectiveness of the proposed method.

关键词： Multi-Focus Image Fusion Discrete Wavelet Transform deep convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

Ethiopian sign language recognition using deep convolutional neural network

引用

MULTIMEDIA TOOLS AND APPLICATIONS 2022年第20期81卷 29027-29043页

作者： Abeje, Bekalu Tadele Salau, Ayodeji Olalekan Mengistu, Abreham Debasu Tamiru, Nigus Kefyalew Haramaya Univ Dept Informat Technol Coll Comp & Informat Dire Dawa Ethiopia Afe Babalola Univ Dept Elect Elect & Comp Engn Ado Ekiti Nigeria Saveetha Inst Med & Tech Sci Saveetha Sch Engn Chennai Tamil Nadu India Bahir Dar Univ Inst Technol Dept Comp Sci Bahir Dar Ethiopia Debre Markos Univ Inst Technol Sch Elect & Comp Engn Debre Markos Ethiopia

In recent years, several technologies have been utilized to bridge the communication gap between persons who have hearing or speaking impairments and those who don't. This paper presents the development of a novel sign language recognition system which translates Ethiopian sign language (ETHSL) to Amharic alphabets using computer vision technology and deep convolutional neural network (CNN). The system accepts sign language images as input and gives Amharic text as the desired output. The proposed system comprises of three main stages which are: preprocessing, feature extraction, and recognition. The methodology employed involves data acquisition, preprocessing the acquired data, background normalization, image resizing, region of interest (ROI) identification, noise removal, brightness adjustment, and feature extraction, while deep convolutional neural network (CNN) was used for end-to-end classification. The data used in this study was acquired from students with hearing impairments at the Debre Markos Teaching College with an iPhone 6s phone which has a resolution of 3024 x 4020. The images are in JPEG file format and were collected in a controlled environment. The proposed system was implemented using Kera's (Tensorflow2.3.0 as backend) in python and tested using the image dataset collected from Debre Markos Teaching College graduating students of 2012. The results show that the running time was minimized by adjusting the images to a suitable size and color. In addition, the results show an improved recognition accuracy compared to previous works. The proposed model achieves 98.5% training, 95.59% validation, and 98.3% testing accuracy of recognition.

关键词： Ethiopian sign language deep convolutional neural network Amharic alphabet

来源：评论

学校读者我要写书评

暂无评论

E²BoWs: An end-to-end Bag-of-Words model via deep convolutional neural network for image retrieval

引用

NEUROCOMPUTING 2020年 395卷 188-198页

作者： Liu, Xiaobin Zhang, Shiliang Huang, Tiejun Tian, Qi Peking Univ Beijing Peoples R China Univ Texas San Antonio Dept Comp Sci San Antonio TX USA

Traditional Bag-of-Words (BoWs) model is commonly generated with many steps, including local feature extraction, codebook generation and feature quantization, etc. Those steps are relatively independent with each other and are hard to be jointly optimized. Moreover, the dependency on hand-crafted local feature makes BoWs model not effective in conveying high-level semantics. These issues largely hinder the performance of BoWs model in large-scale image applications. To conquer these issues, we propose an End-to-End BoWs (E(2)BoWs) model based on deep convolutional neural network (DCNN). Our model takes an image as input, then identifies and separates semantic objects in it, and finally outputs visual words with high semantic discriminative power. Specifically, our model firstly generates Semantic Feature Maps (SFMs) corresponding to different object categories through convolutional layers, then introduces Bag-of-Words Layers (BoWL) to generate visual words from each individual feature map. We also introduce a novel learning algorithm to reinforce the sparsity of the generated E(2)BoWs model, which further ensures the time and memory efficiency. We evaluate the proposed E(2)BoWs model on several image search datasets including MNIST, SVHN, CIFAR-10, CIFAR-100, MIRFLICKR-25K and NUS-WIDE. Experimental results show that our method achieves promising accuracy and efficiency compared with recent deep learning based retrieval works. (C) 2019 Published by Elsevier B.V.

关键词： Large-scale image retrieval Bag-of-Words Model deep convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

A Combination Model of Shifting Joint Angle Changes With 3D-deep convolutional neural network to Recognize Human Activity

引用

IEEE TRANSACTIONS ON neural SYSTEMS AND REHABILITATION ENGINEERING 2024年 32卷 1078-1089页

作者： Rahayu, Endang Sri Yuniarno, Eko Mulyanto Purnama, I. Ketut Eddy Purnomo, Mauridhi Hery Inst Teknol Sepuluh Nopember Dept Elect Engn Surabaya 60111 Indonesia Univ Jayabaya Dept Elect Engn Jakarta 13210 Indonesia Inst Teknol Sepuluh Nopember Dept Comp Engn Surabaya 60111 Indonesia

Research in the field of human activity recognition is very interesting due to its potential for various applications such as in the field of medical rehabilitation. The need to advance its development has become increasingly necessary to enable efficient detection and response to a wide range of movements. Current recognition methods rely on calculating changes in joint distance to classify activity patterns. Therefore, a different approach is required to identify the direction of movement to distinguish activities exhibiting similar joint distance changes but differing motion directions, such as sitting and standing. The research conducted in this study focused on determining the direction of movement using an innovative joint angle shift approach. By analyzing the joint angle shift value between specific joints and reference points in the sequence of activity frames, the research enabled the detection of variations in activity direction. The joint angle shift method was combined with a deep convolutional neural network (DCNN) model to classify 3D datasets encompassing spatial-temporal information from RGB-D video image data. Model performance was evaluated using the confusion matrix. The results show that the model successfully classified nine activities in the Florence 3D Actions dataset, including sitting and standing, obtaining an accuracy of (96.72 +/- 0.83)%. In addition, to evaluate its robustness, this model was tested on the UTKinect Action3D dataset, obtaining an accuracy of 97.44%, proving that state-of-the-art performance has been achieved.

关键词： Combination model deep convolutional neural network human activity recognition shifting joint angles

来源：评论

学校读者我要写书评

暂无评论

Human Activity Recognition Based On Video Summarization And deep convolutional neural network

引用

COMPUTER JOURNAL 2024年第8期67卷 2601-2609页

作者： Kushwaha, Arati Khare, Manish Bommisetty, Reddy Mounika Khare, Ashish GLA Univ Dept Comp Engn & Applicat Mathura India Dhirubhai Ambani Inst Informat & Commun Technol Gandhinagar India Univ Allahabad Dept Elect & Commun Prayagraj Uttar Pradesh India

In this technological era, human activity recognition (HAR) plays a significant role in several applications like surveillance, health services, Internet of Things, etc. Recent advancements in deep learning and video summarization have motivated us to integrate these techniques for HAR. This paper introduces a computationally efficient HAR technique based on a deep learning framework, which works well in realistic and multi-view environments. deep convolutional neural networks (DCNNs) normally suffer from different constraints, including data size dependencies, computational complexity, overfitting, training challenges and vanishing gradients. Additionally, with the use of advanced mobile vision devices, the demand for computationally efficient HAR algorithms with the requirement of limited computational resources is high. To address these issues, we used integration of DCNN with video summarization using keyframes. The proposed technique offers a solution that enhances performance with efficient resource utilization. For this, first, we designed a lightweight and computationally efficient deep learning architecture based on the concept of identity skip connections (features reusability), which preserves the gradient loss attenuation and can handle the enormous complexity of activity classes. Subsequently, we employed an efficient keyframe extraction technique to minimize redundancy and succinctly encapsulate the entire video content in a lesser number of frames. To evaluate the efficacy of the proposed method, we performed the experimentation on several publicly available datasets. The performance of the proposed method is measured in terms of evaluation parameters Precision, Recall, F-Measure and Classification Accuracy. The experimental results demonstrated the superiority of the presented algorithm over other existing state-of-the-art methods.

关键词： human activity recognition deep convolutional neural network video summarization keyframes feature reusability identity skip connections softmax classifier

来源：评论

学校读者我要写书评

暂无评论

Ultrasonographic Thyroid Nodule Classification Using a deep convolutional neural network with Surgical Pathology

引用

JOURNAL OF DIGITAL IMAGING 2020年第5期33卷 1202-1208页

作者： Kwon, Soon Woo Choi, Ik Joon Kang, Ju Yong Jang, Won Il Lee, Guk-Haeng Lee, Myung-Chul Korea Inst Radiol & Med Sci KIRAMS Radiat Med Clin Res Div Seoul South Korea Korea Inst Radiol & Med Sci KIRAMS Dept Otorhinolaryngol Korea Canc Ctr Hosp 75 Nowon Gil Seoul 139706 South Korea Korea Inst Radiol & Med Sci KIRAMS Korea Canc Ctr Hosp Radiat Oncol Seoul South Korea

Ultrasonography with fine-needle aspiration biopsy is commonly used to detect thyroid cancer. However, thyroid ultrasonography is prone to subjective interpretations and interobserver variabilities. The objective of this study was to develop a thyroid nodule classification system for ultrasonography using convolutional neural networks. Transverse and longitudinal ultrasonographic thyroid images of 762 patients were used to create a deep learning model. After surgical biopsy, 325 cases were confirmed to be benign and 437 cases were confirmed to be papillary thyroid carcinoma. Image annotation marks were removed, and missing regions were recovered using neighboring parenchyme. To reduce overfitting of the deep learning model, we applied data augmentation, global average pooling. And 4-fold cross-validation was performed to detect overfitting. We employed a transfer learning method with the pretrained deep learning model VGG16. The average area under the curve of the model was 0.916, and its specificity and sensitivity were 0.70 and 0.92, respectively. Positive and negative predictive values were 0.90 and 0.75, respectively. We introduced a new fine-tuned deep learning model for classifying thyroid nodules in ultrasonography. We expect that this model will help physicians diagnose thyroid nodules with ultrasonography.

关键词： deep convolutional neural network deep learning Ultrasonography Thyroid nodule classification

来源：评论

学校读者我要写书评

暂无评论

A novel adversarial transfer learning in deep convolutional neural network for intelligent diagnosis of gas-insulated switchgear insulation defect A DATCNN for GIS insulation defect diagnosis

引用

IET GENERATION TRANSMISSION & DISTRIBUTION 2021年第23期15卷 3229-3241页

作者： Wang, Yanxin Yan, Jing Jing, Qianzhen Qi, Zhenkang Wang, Jianhua Geng, Yingsan Xi An Jiao Tong Univ Stare Key Lab Elect Insulat & Power Equip Xian 710049 Shaanxi Peoples R China Tsinghua Univ State Key Lab Power Syst Beijing 100084 Peoples R China

Recently, numerous data-driven fault diagnosis methods have been developed, and the tasks involving the same distribution of training and test data have been well solved. However, considering the particularity of gas-insulated switchgear (GIS), collecting massive data, especially with the same distribution, is difficult. Therefore, existing fault diagnosis methods hardly achieve satisfactory insulation defect diagnosis with small datasets. Aiming at solving this problem, a novel domain adversarial transfer convolutional neural network (DATCNN) is proposed, realising the diagnosis of GIS insulation defects on small samples. First, a residual CNN is built to learn feature representations from the source and target domains. Second, the domain adversarial training strategy is used for feature transfer, where a conditional adversarial mechanism is introduced, and the joint distribution of features and labels is improved to a random linear combination, which realises the simultaneous adaptation of features and labels. Finally, the Nesterov accelerated gradient descent optimisation algorithm is used to speed up the gradient convergence. DATCNN has 99.15% and >= 89.5% diagnosis accuracy for GIS insulation defects in the laboratory and on-site, respectively. Comprehensive experiment results show the effectiveness and superiority of the proposed method in diagnosing GIS insulation defects with small samples.

关键词： gradient convergence fault diagnosis data-driven fault diagnosis residual CNN Optimisation techniques domain adversarial transfer convolutional neural network feature representations deep convolutional neural network intelligent diagnosis conditional adversarial mechanism convergence of numerical methods adversarial transfer learning random linear combination convolutional neural nets deep learning (artificial intelligence) gas-insulated switchgear insulation defect gas insulated switchgear optimisation Interpolation and function approximation (numerical analysis) GIS insulation defect diagnosis DATCNN power engineering computing domain adversarial training Power engineering computing Switchgear feature transfer Nesterov accelerated gradient descent optimisation gradient methods

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：