scholarly journals Implementasi Metode Convolutional Neural Network Menggunakan Arsitektur LeNet-5 untuk Pengenalan Doodle

2020 ◽  
Vol 1 (1) ◽  
pp. 45-56
Author(s):  
Muhammad Rafly Alwanda ◽  
Raden Putra Kurniawan Ramadhan ◽  
Derry Alamsyah
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Test Results ◽  
The Third ◽  
Recognition Of Objects ◽  
Handwritten Recognition

Recognition of objects to date has been widely applied in various fields, for example in handwritten recognition. This research utilizes the ability of CNN to use LeNet-5 architecture for the introduction of doodle types with 5 object images, namely clothes, pants, chairs, butterflies and bicycles. Each doodle object consists of 30 images with a total dataset of 150 images. The test results show that the first, second and fourth scenarios of bicycle objects are more recognized with an accuracy value of 93% - 98%, recall 86% - 93% and precision 81% - 93%, clothes objects are more recognized in the third scenario with an accuracy value of 94%, 86% recall, and 83% precision.

Research on Computer-Aided Diagnosis of Alzheimer’s Disease Based on Heterogeneous Medical Data Fusion

2019 ◽  
Vol 33 (05) ◽  
pp. 1957001
Author(s):  
Yin Dai ◽  
Daoyun Qiu ◽  
Yang Wang ◽  
Sizhe Dong ◽  
Hong-Li Wang
Keyword(s):  
Neural Network ◽  
Alzheimer’S Disease ◽  
Alzheimer's Disease ◽  
Convolutional Neural Network ◽  
Effective Means ◽  
Computer Aided Diagnosis ◽  
Test Results ◽  
Back Propagation Algorithm ◽  
Computer Aided ◽  
Aided Diagnosis

Alzheimer’s disease is the third most expensive disease, only after cancer and cardiopathy. It is also the fourth leading cause of death in the elderly after cardiopathy, cancer, and cerebral palsy. The disease lacks specific diagnostic criteria. At present, there is still no definitive and effective means for preclinical diagnosis and treatment. It is the only disease that cannot be prevented and cured among the world’s top ten fatal diseases. It has now been proposed as a global issue. Computer-aided diagnosis of Alzheimer’s disease (AD) is mostly based on images at this stage. This project uses multi-modality imaging MRI/PET combining with clinical scales and uses deep learning-based computer-aided diagnosis to treat AD, improves the comprehensiveness and accuracy of diagnosis. The project uses Bayesian model and convolutional neural network to train experimental data. The experiment uses the improved existing network model, LeNet-5, to design and build a 10-layer convolutional neural network. The network uses a back-propagation algorithm based on a gradient descent strategy to achieve good diagnostic results. Through the calculation of sensitivity, specificity and accuracy, the test results were evaluated, good test results were obtained.

scholarly journals Image-based failure detection for material extrusion process using a convolutional neural network

2020 ◽  
Vol 111 (5-6) ◽  
pp. 1291-1302
Author(s):  
Hyungjung Kim ◽  
Hyunsu Lee ◽  
Ji-Soo Kim ◽  
Sung-Hoon Ahn
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Failure Detection ◽  
Extrusion Process ◽  
Shape Error ◽  
Material Extrusion ◽  
Recognition Of Objects ◽  
Deposition Processes ◽  
Printing Processes ◽  
Visual Geometry

Abstract The material extrusion (ME) process is one of the most widely used 3D printing processes, especially considering its use of inexpensive materials. However, the error known as the “spaghetti-shape error,” related to filament tangling, is a common problem associated with the ME process. Once occurring, this issue, which consumes both time and materials, requires a restart of the entire process. In order to prevent this, the user must constantly monitor the process. In this research, a failure detection method which uses a webcam and deep learning is developed for the ME process. The webcam captures images and then analyzes them by machine learning based on a convolutional neural network (CNN), showing outstanding performance in both image classification and the recognition of objects. Sample images were trained based on a modified Visual Geometry Group Network (VGGNet) model and the trained model was evaluated, resulting in 97% accuracy. The pre-trained model was tested on a 3D printer monitoring system for its ability to recognize the “spaghetti-shape-error” and was able to detect 96% of abnormal deposition processes. The proposed method can analyze the ME process in real time and informs the user or halts the process when abnormal printing is detected.

scholarly journals A Method of Containment Leakage Rate Estimation Based on Convolution Neural Network

2021 ◽  
Vol 9 ◽  
Author(s):  
Hao Wang ◽  
Jingquan Liu ◽  
Guangyao Xie ◽  
Xianping Zhong ◽  
Xiangqi Fan
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Nuclear Power ◽  
Power Plants ◽  
Least Square Method ◽  
Leakage Rate ◽  
Least Square ◽  
Nonlinear Fitting ◽  
The Third ◽  
Start Up

As the nuclear power plant containment is the third barrier to nuclear safety, real-time monitoring of containment leakage rate is very important in addition to the overall leakage test before an operation. At present, most of the containment leakage rate monitoring systems calculate the standard volume of moist air in the containment through monitoring parameters and calculate the daily leakage rate by the least square method. This method requires several days of data accumulation to accurately calculate. In this article, a new leakage rate modeling technique is proposed using a convolutional neural network based on data of the monitoring system. Use the daily monitoring parameters of nuclear power plants to construct inputs of the model and train the convolutional neural network with daily leakage rates as labels. This model makes use of the powerful nonlinear fitting ability of the convolutional neural network. It can use 1-day data to accurately calculate the containment leakage rate during the reactor start-up phase and can timely determine whether the containment leak has occurred during the start-up phase and deal with it in time, to ensure the integrity of the third barrier.

scholarly journals Message Insertion Using the Convolutional Neural Network Model Approach

SinkrOn ◽  
2019 ◽  
Vol 4 (1) ◽  
pp. 215
Author(s):  
Lita Ambarwati ◽  
Agrifa Darwanto Sirait ◽  
Bella Siska Tambun ◽  
Eko Paskah Jeremia Purwanto ◽  
Amir Mahmud Husein
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Computing Time ◽  
Cover Image ◽  
Test Results ◽  
Model Framework ◽  
Secret Image ◽  
Lighting Conditions ◽  
Model Approach

One problem in computer vision that has long been sought for a solution is the classification of objects in the image in general. How to duplicate the ability of humans to understand image information, so that computers can recognize objects in the image as humans do. The feature engineering process used is generally very limited where it can only apply to certain datasets without the ability to generalize to any type of image. That is because various differences between images include differences in perspective, differences in scale, differences in lighting conditions, deformation of objects, and so on. Academics who have long struggled with this issue. The application of the Convolutional Neural Network (CNN) method for the insertion of messages in an image with the aim of securing the proposed message produces good security, from the test results, it can be concluded as follows The Convolutional Neural Network (CNN) method requires computing time to insert messages in a secret image. The model framework uses 2 (two) images with the aim of the cover image as input and the secret image where the secret image has been inserted a message so that the secret is not visible. The cover image that has been inserted a secret picture that contains the message looks not much different, but the file size of the secret picture has increased by 66%.

scholarly journals Deteksi Parasit Plasmodium pada Citra Mikroskopis Hapusan Darah dengan Metode Deep Learning

2021 ◽  
Vol 9 (2) ◽  
pp. 306
Author(s):  
NOR KUMALASARI CAECAR PRATIWI ◽  
NUR IBRAHIM ◽  
YUNENDAH NUR FU’ADAH ◽  
SYAMSUL RIZAL
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Convolutional Neural Network ◽  
Anopheles Mosquito ◽  
Microscopic Image ◽  
Test Results ◽  
Automatic Diagnosis ◽  
Diagnosis System ◽  
Slowing Down ◽  
The World

ABSTRAKParasit plasmodium merupakan makhluk protozoa bersel satu yang menjadi penyebab penyakit malaria. Plasmodium ini dibawa melalui gigitan nyamuk anopheles betina. Dalam World Malaria Report 2015 menyatakan bahwa malaria telah menyerang sedikit 106 negara di dunia. Di Indonesia sendiri, Papua, NTT dan Maluku merupakan wilayah dengan kasus positif malaria tertinggi. Malaria telah menjadi masalah yang serius, sehingga keberadaan sistem diagnosa otomatis yang cepat dan handal sangat diperlukan untuk proses perlambatan penyeberan dan pembasmian epidemi. Dalam penelitian ini akan dirancang sistem yang mampu mendeteksi parasit malaria pada citra mikroskopis darah menggunakan arsitekur Convolutional Neural Network (CNN) sederhana. Hasil pengujian menunjukkan bahwa metode yang diusulkan memberikan presisi dan recall sebesar 0,98 dan f1-score sebesar 0,96 serta akurasi 95,83%.Kata kunci: parasit, malaria, convolutional neural network, citra mikroskopis ABSTRACTPlasmodium parasites are single-celled protozoan creatures that cause malaria. Plasmodium is carried through the bite of a female Anopheles mosquito. The World Malaria Report 2015 states that malaria has attacked at least 106 countries in the world. In Indonesia itself, Papua, NTT and Maluku are the regions with the highest positive cases of malaria. Malaria has become a serious problem, so the existence of a fast and reliable automatic diagnosis system is indispensable for the process of slowing down the spread and eliminating the epidemic. In this study, a system capable of detecting malaria parasites in microscopic images of blood will be designed using a simple Convolutional Neural Network (CNN) architecture. The test results show that the proposed method provides precision and recall of 0,98, f1-values of 0.96 and accuracy of 95,83%.Keywords: parasites, malaria, convolutional neural network, microscopic image

scholarly journals Application of Deep Learning Convolution Neural Network Method on KRSBI Humanoid R-SCUAD Robot

2020 ◽  
Vol 2 (1) ◽  
pp. 40
Author(s):  
Syahid Al Irfan ◽  
Nuryono Satya Widodo
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Real Time ◽  
Architectural Design ◽  
Humanoid Robots ◽  
White Ball ◽  
Test Results ◽  
Object Classes ◽  
Hidden Layer ◽  
Ball Detection

In a soccer game the ability of humanoid robots that one needs to have is to see the ball object in real time. Development of the ability of humanoid robots to see the ball has been developed but the level of accuracy of object recognition and adaptation during matches still needs to be improved. The architecture designed in this study is Convolutional Neural Network or CNN which is designed to have 6 hidden layers with implementation of the robot program using the Tensorflow library. The pictures taken are used in the training process to have 9 types of images based on where the pictures were taken. Each type of image is divided into 2 classes, namely 2000 images for ball object classes and 2000 images for non-ball object classes. The test is done in real time using a white ball on green grass. From the architectural design and white ball detection test results obtained a success rate of 67%, five of the nine models managed to recognize the ball. The model can recognize objects with an image processing speed of a maximum of 13 FPS.Dalam pertandingan sepak bola kemampuan robot humanoid yang perlu dimiliki salah satunya adalah melihat objek bola secara real time. Pengembangan kemampuan robot humanoid untuk melihat bola telah dikembangkan tetapi tingkat akurasi pengenalan objek dan adaptasi saat pertandingan masih perlu ditingkatkan. Arsitektur yang dirancang pada penelitian ini yaitu Convolutional Neural Network atau CNN yang dirancang memiliki 6 hidden layer dengan implementasi pada program robot menggunakan library Tensorflow. Gambar yang diambil digunakan dalam proses training memiliki 9 jenis gambar berdasarkan tempat pengambilan gambar. Tiap jenis gambar terbagi menjadi 2 class yaitu 2000 gambar untuk class objek bola dan 2000 gambar untuk class objek bukan bola. Pengujian dilakukan secara real time dengan menggunakan bola berwarna putih di atas rumput hijau. Dari perancangan arsitektur dan hasil pengujian pendeteksian bola putih didapatkan persentase keberhasilan 67% yaitu lima dari sembilan model berhasil mengenali bola. Model dapat mengenali objek dengan kecepatan pengolahan gambar adalah maksimal 13 FPS.

scholarly journals CNID: Research of Network Intrusion Detection Based on Convolutional Neural Network

2020 ◽  
Vol 2020 ◽  
pp. 1-11
Author(s):  
Guojie Liu ◽  
Jianbiao Zhang
Keyword(s):  
Neural Network ◽  
Intrusion Detection ◽  
Convolutional Neural Network ◽  
Detection System ◽  
False Positive Rate ◽  
Network Intrusion Detection ◽  
Test Results ◽  
Softmax Classifier ◽  
Detection Model ◽  
Network Intrusion

Network intrusion detection system can effectively detect network attack behaviour, which is very important to network security. In this paper, a multiclassification network intrusion detection model based on convolutional neural network is proposed, and the algorithm is optimized. First, the data is preprocessed, the original one-dimensional network intrusion data is converted into two-dimensional data, and then the effective features are learned using optimized convolutional neural networks, and, finally, the final test results are produced in conjunction with the Softmax classifier. In this paper, KDD-CUP 99 and NSL-KDD standard network intrusion detection dataset were used to carry out the multiclassification network intrusion detection experiment; the experimental results show that the multiclassification network intrusion detection model proposed in this paper improves the accuracy and check rate, reduces the false positive rate, and also obtains better test results for the detection of unknown attacks.

scholarly journals EkushNet: Using Convolutional Neural Network for Bangla Handwritten Recognition

2018 ◽  
Vol 143 ◽  
pp. 603-610 ◽  
Author(s):  
AKM Shahariar Azad Rabby ◽  
Sadeka Haque ◽  
Sheikh Abujar ◽  
Syed Akhter Hossain
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Handwritten Recognition
Sign in / Sign up

Export Citation Format

Share Document