scholarly journals Skin Lesion Classification Using Deep Learning Methods

2020 ◽  
Vol 15 (2) ◽  
pp. 180-194
Author(s):  
E.Yu. Shchetinin ◽  
A.V. Demidova ◽  
D.S. Kulyabov ◽  
L.A. Sevastyanov
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Skin Lesion ◽  
Deep Neural Network ◽  
Computer Experiments ◽  
Test Sample ◽  
Skin Lesions ◽  
Learning Methods ◽  
Data Set ◽  
Noise Contamination

In this paper, we propose an approach to solving the problem of recognizing skin lesions, namely melanoma, based on the analysis of dermoscopic images using deep learning methods. For this purpose, the architecture of a deep convolutional neural network was developed, which was applied to the processing of dermoscopic images of various skin lesions contained in the HAM10000 data set. The data under study were preprocessed to eliminate noise, contamination, and change the size and format of images. In addition, since the disease classes are unbalanced, a number of transformations were performed to balance them. The data obtained in this way were divided into two classes: Melanoma and Benign. Computer experiments using the built deep neural network based on the data obtained in this way have shown that the proposed approach provides 94% accuracy on the test sample, which exceeds similar results obtained by other algorithms.

scholarly journals RISC-V Virtual Platform-Based Convolutional Neural Network Accelerator Implemented in SystemC

Electronics ◽  
2021 ◽  
Vol 10 (13) ◽  
pp. 1514
Author(s):  
Seung-Ho Lim ◽  
WoonSik William Suh ◽  
Jin-Young Kim ◽  
Sang-Young Cho
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Network Model ◽  
Neural Network Model ◽  
Deep Neural Network ◽  
System Level ◽  
Neural Network Models ◽  
Data Set ◽  
Embedded Device ◽  
Virtual Platform

The optimization for hardware processor and system for performing deep learning operations such as Convolutional Neural Networks (CNN) in resource limited embedded devices are recent active research area. In order to perform an optimized deep neural network model using the limited computational unit and memory of an embedded device, it is necessary to quickly apply various configurations of hardware modules to various deep neural network models and find the optimal combination. The Electronic System Level (ESL) Simulator based on SystemC is very useful for rapid hardware modeling and verification. In this paper, we designed and implemented a Deep Learning Accelerator (DLA) that performs Deep Neural Network (DNN) operation based on the RISC-V Virtual Platform implemented in SystemC in order to enable rapid and diverse analysis of deep learning operations in an embedded device based on the RISC-V processor, which is a recently emerging embedded processor. The developed RISC-V based DLA prototype can analyze the hardware requirements according to the CNN data set through the configuration of the CNN DLA architecture, and it is possible to run RISC-V compiled software on the platform, can perform a real neural network model like Darknet. We performed the Darknet CNN model on the developed DLA prototype, and confirmed that computational overhead and inference errors can be analyzed with the DLA prototype developed by analyzing the DLA architecture for various data sets.

scholarly journals Dynamic Wide and Deep Neural Network for Hyperspectral Image Classification

2021 ◽  
Vol 13 (13) ◽  
pp. 2575
Author(s):  
Jiangbo Xi ◽  
Ming Cong ◽  
Okan K. Ersoy ◽  
Weibao Zou ◽  
Chaoying Zhao ◽  
...  
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Deep Neural Network ◽  
Hyperspectral Image ◽  
Least Squares Method ◽  
Support Vector ◽  
Learning Networks ◽  
Learning Methods ◽  
Sliding Windows ◽  
Deep Layers

Recently, deep learning has been successfully and widely used in hyperspectral image (HSI) classification. Considering the difficulty of acquiring HSIs, there are usually a small number of pixels used as the training instances. Therefore, it is hard to fully use the advantages of deep learning networks; for example, the very deep layers with a large number of parameters lead to overfitting. This paper proposed a dynamic wide and deep neural network (DWDNN) for HSI classification, which includes multiple efficient wide sliding window and subsampling (EWSWS) networks and can grow dynamically according to the complexity of the problems. The EWSWS network in the DWDNN was designed both in the wide and deep direction with transform kernels as hidden units. These multiple layers of kernels can extract features from the low to high level, and because they are extended in the wide direction, they can learn features more steadily and smoothly. The sliding windows with the stride and subsampling were designed to reduce the dimension of the features for each layer; therefore, the computational load was reduced. Finally, all the weights were only from the fully connected layer, and the iterative least squares method was used to compute them easily. The proposed DWDNN was tested with several HSI data including the Botswana, Pavia University, and Salinas remote sensing datasets with different numbers of instances (from small to big). The experimental results showed that the proposed method had the highest test accuracies compared to both the typical machine learning methods such as support vector machine (SVM), multilayer perceptron (MLP), radial basis function (RBF), and the recently proposed deep learning methods including the 2D convolutional neural network (CNN) and the 3D CNN designed for HSI classification.

scholarly journals Tri-net for Semi-Supervised Deep Learning

2018 ◽  
Author(s):  
Dong-Dong Chen ◽  
Wei Wang ◽  
Wei Gao ◽  
Zhi-Hua Zhou
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Learning ◽  
Error Rate ◽  
Deep Neural Network ◽  
Deep Neural Networks ◽  
State Of The Art ◽  
Fine Tuning ◽  
Learning Methods ◽  
Model Initialization

Deep neural networks have witnessed great successes in various real applications, but it requires a large number of labeled data for training. In this paper, we propose tri-net, a deep neural network which is able to use massive unlabeled data to help learning with limited labeled data. We consider model initialization, diversity augmentation and pseudo-label editing simultaneously. In our work, we utilize output smearing to initialize modules, use fine-tuning on labeled data to augment diversity and eliminate unstable pseudo-labels to alleviate the influence of suspicious pseudo-labeled data. Experiments show that our method achieves the best performance in comparison with state-of-the-art semi-supervised deep learning methods. In particular, it achieves 8.30% error rate on CIFAR-10 by using only 4000 labeled examples.

scholarly journals Skin Lesion Segmentation Using Local Binary Convolution-Deconvolution Architecture

2020 ◽  
Vol 39 (3) ◽  
pp. 169-185
Author(s):  
Omran Salih ◽  
Serestina Viriri
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Skin Lesion ◽  
Convolutional Neural Network ◽  
Skin Lesions ◽  
Training Dataset ◽  
Great Success ◽  
Lesion Segmentation ◽  
Convolutional Network ◽  
Low Contrast

Deep learning techniques such as Deep Convolutional Networks have achieved great success in skin lesion segmentation towards melanoma detection. The performance is however restrained by distinctive and challenging features of skin lesions such as irregular and fuzzy border, noise and artefacts presence and low contrast between lesions. The methods are also restricted with scarcity of annotated lesion images training dataset and limited computing resources. Recent research in convolutional neural network (CNN) has provided a variety of new architectures for deep learning. One interesting new architecture is the local binary convolutional neural network (LBCNN), which can reduce the workload of CNNs and improve the classification accuracy. The proposed framework employs the local binary convolution on U-net architecture instead of the standard convolution in order to reduced-size deep convolutional encoder-decoder network that adopts loss function for robust segmentation. The proposed framework replaced the encoder part in U-net by LBCNN layers. The approach automatically learns and segments complex features of skin lesion images. The encoder stage learns the contextual information by extracting discriminative features while the decoder stage captures the lesion boundaries of the skin images. This addresses the issues with encoder-decoder network producing coarse segmented output with challenging skin lesions appearances such as low contrast between healthy and unhealthy tissues and fine grained variability. It also addresses issues with multi-size, multi-scale and multi-resolution skin lesion images. The deep convolutional network also adopts a reduced-size network with just five levels of encoding-decoding network. This reduces greatly the consumption of computational processing resources. The system was evaluated on publicly available dataset of ISIC and PH2. The proposed system outperforms most of the existing state-of-art.

scholarly journals Deep learning the holographic black hole with charge

2019 ◽  
Vol 28 (12) ◽  
pp. 1950153 ◽  
Author(s):  
Jing Tan ◽  
Chong-Bin Chen
Keyword(s):  
Neural Network ◽  
Black Hole ◽  
Deep Learning ◽  
Deep Neural Network ◽  
Learning Algorithm ◽  
Equation Of Motion ◽  
Batch Size ◽  
Training Process ◽  
Data Set ◽  
Deep Learning Algorithm

We use the deep learning algorithm to learn the Reissner–Nordström (RN) black hole metric by building a deep neural network. Plenty of data are determined in boundary of AdS and we propagate them to the black hole horizon through AdS metric and equation of motion (e.o.m). We label these data according to the values near the horizon, and together with initial data they constitute a data set. Then we construct corresponding deep neural network and train it with the data set to obtain the Reissner–Nordström (RN) black hole metric. Finally, we discuss the effects of learning rate, batch-size and initialization on the training process.

scholarly journals Classification of Diabetic Retinopathy using Stacked Autoencoder-Based Deep Neural Network

2020 ◽  
Vol 1 (1) ◽  
pp. 9-14
Author(s):  
Yasir Eltigani Ali Mustaf ◽  
◽  
Bashir Hassan Ismail ◽  
Keyword(s):  
Neural Network ◽  
Diabetic Retinopathy ◽  
Deep Learning ◽  
Deep Neural Network ◽  
Data Set ◽  
Learning Framework ◽  
Training Images ◽  
Stacked Autoencoder

Diagnosis of diabetic retinopathy (DR) via images of colour fundus requires experienced clinicians to determine the presence and importance of a large number of small characteristics. This work proposes and named Adapted Stacked Auto Encoder (ASAE-DNN) a novel deep learning framework for diabetic retinopathy (DR), three hidden layers have been used to extract features and classify them then use a Softmax classification. The models proposed are checked on Messidor's data set, including 800 training images and 150 test images. Exactness, accuracy, time, recall and calculation are assessed for the outcomes of the proposed models. The results of these studies show that the model ASAE-DNN was 97% accurate.

scholarly journals Hybrid CNN-LSTM Model for Answer Identification

2019 ◽  
Vol 8 (3) ◽  
pp. 1163-1166
Keyword(s):  
Neural Network ◽  
Feature Extraction ◽  
Deep Learning ◽  
Comparative Analysis ◽  
Hybrid Model ◽  
Deep Neural Network ◽  
Extraction Methods ◽  
Learning Methods ◽  
Automatic Feature Extraction

User quest for information has led to development of Question Answer (QA) system to provide relevant answers to user questions. The QA task are different than normal NLP tasks as they heavily depend to semantics and context of given data. Retrieving and predicting answers to verity of questions require understanding of question, relevance with context and identifying and retrieving of suitable answers. Deep learning helps to produce impressive performance as it employs deep neural network with automatic feature extraction methods. The paper proposes a hybrid model to identify suitable answer for posed question. The proposes power exploits the power of CNN for extracting features and ability of LSTM for considering long term dependencies and semantic of context and question. Paper provides a comparative analysis on deep learning methods useful for predicting answer with the proposed method .The model is implemented on twenty tasks of babI dataset of Facebook .

scholarly journals A Model of Convolutional Neural Network Combined with External Knowledge to Measure the Question Similarity for Community Question Answering Systems

2021 ◽  
Vol 11 (3) ◽  
pp. 194-201
Author(s):  
Van-Tu Nguyen ◽  
◽  
Anh-Cuong Le ◽  
Ha-Nam Nguyen
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Convolutional Neural Network ◽  
Question Answering ◽  
Learning Models ◽  
Learning Methods ◽  
Data Set ◽  
External Knowledge ◽  
Community Question Answering ◽  
Proposed Model

Automatically determining similar questions and ranking the obtained questions according to their similarities to each input question is a very important task to any community Question Answering system (cQA). Various methods have applied for this task including conventional machine learning methods with feature extraction and some recent studies using deep learning methods. This paper addresses the problem of how to combine advantages of different methods into one unified model. Moreover, deep learning models are usually only effective for large data, while training data sets in cQA problems are often small, so the idea of integrating external knowledge into deep learning models for this cQA problem becomes more important. To this objective, we propose a neural network-based model which combines a Convolutional Neural Network (CNN) with features from other methods so that the deep learning model is enhanced with addtional knowledge sources. In our proposed model, the CNN component will learn the representation of two given questions, then combined additional features through a Multilayer Perceptron (MLP) to measure similarity between the two questions. We tested our proposed model on the SemEval 2016 task-3 data set and obtain better results in comparison with previous studies on the same task.

scholarly journals Research on Time Series Anomaly Detection: Based on Deep Learning Methods

2021 ◽  
Vol 2132 (1) ◽  
pp. 012012
Author(s):  
Jiaqi Zhou
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Time Series ◽  
Deep Learning ◽  
Anomaly Detection ◽  
Deep Neural Network ◽  
Deep Neural Networks ◽  
Detection Task ◽  
Detection Methods ◽  
Learning Methods

Abstract Time series anomaly detection has always been an important research direction. The early time series anomaly detection methods are mainly statistical methods and machine learning methods. With the powerful functions of deep neural network being continuously mined by researchers, the effect of deep neural network in anomaly detection task has been significantly better than the traditional methods. In view of the continuous development and application of deep neural networks such as transformer and graph neural network (GNN) in time series anomaly detection in recent years, the body of research lacks a comparative evaluation of deep learning methods in recent years. This paper studies various deep neural networks suitable for time series, which are divided into three categories according to anomaly detection methods. The evaluation is conducted on public datasets. By analyzing the evaluation criteria, this paper discusses the performance of each model, as well as the problems and development direction in the field of time series anomaly detection in the future. This study found that in the time series anomaly detection task, transformer is suitable for dealing with long-time series prediction, and studying the graph structure of time series may be the best way to deal with time series anomaly detection in the future

scholarly journals Design of a Broadband Solar Thermal Absorber Using a Deep Neural Network and Experimental Demonstration of Its Performance

2019 ◽  
Vol 9 (1) ◽  
Author(s):  
Junyong Seo ◽  
Pil-Hoon Jung ◽  
Mingeon Kim ◽  
Sounghyeok Yang ◽  
Ikjin Lee ◽  
...  
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Deep Neural Network ◽  
Solar Spectrum ◽  
Solar Thermal ◽  
Data Set ◽  
Rigorous Coupled Wave Analysis ◽  
Absorption Performance ◽  
Rigorous Coupled Wave ◽  
Computational Resources

Abstract In using nanostructures to design solar thermal absorbers, computational methods, such as rigorous coupled-wave analysis and the finite-difference time-domain method, are often employed to simulate light-structure interactions in the solar spectrum. However, those methods require heavy computational resources and CPU time. In this study, using a state-of-the-art modeling technique, i.e., deep learning, we demonstrate significant reduction of computational costs during the optimization processes. To minimize the number of samples obtained by actual simulation, only regulated amounts are prepared and used as a data set to train the deep neural network (DNN) model. Convergence of the constructed DNN model is carefully examined. Moreover, several analyses utilizing an evolutionary algorithm, which require a remarkable number of performance calculations, are performed using the trained DNN model. We show that deep learning effectively reduces the actual simulation counts compared to the case of a design process without a neural network model. Finally, the proposed solar thermal absorber is fabricated and its absorption performance is characterized.

Sign in / Sign up

Export Citation Format

Share Document