scholarly journals Deep Learning to Detect and Classify the Purity Level of Luwak Coffee Green Beans

2021 ◽  
Vol 30 (1) ◽  
pp. 1-18
Author(s):  
Yusuf Hendrawan ◽  
Shinta Widyaningtyas ◽  
Muchammad Riza Fauzy ◽  
Sucipto Sucipto ◽  
Retno Damayanti ◽  
...  
Keyword(s):  
Deep Learning ◽  
Detection System ◽  
Low Cost ◽  
Confusion Matrix ◽  
Optimization Technique ◽  
Learning Rate ◽  
Production Costs ◽  
Green Beans ◽  
Initial Learning ◽  
Palm Civet

Luwak coffee (palm civet coffee) is known as one of the most expensive coffee in the world. In order to lower production costs, Indonesian producers and retailers often mix high-priced Luwak coffee with regular coffee green beans. However, the absence of tools and methods to classify Luwak coffee counterfeiting makes the sensing method’s development urgent. The research aimed to detect and classify Luwak coffee green beans purity into the following purity categories, very low (0-25%), low (25-50%), medium (50-75%), and high (75-100%). The classifying method relied on a low-cost commercial visible light camera and the deep learning model method. Then, the research also compared the performance of four pre-trained convolutional neural network (CNN) models consisting of SqueezeNet, GoogLeNet, ResNet-50, and AlexNet. At the same time, the sensitivity analysis was performed by setting the CNN parameters such as optimization technique (SGDm, Adam, RMSProp) and the initial learning rate (0.00005 and 0.0001). The training and validation result obtained the GoogLeNet as the best CNN model with optimizer type Adam and learning rate 0.0001, which resulted in 89.65% accuracy. Furthermore, the testing process using confusion matrix from different sample data obtained the best CNN model using ResNet-50 with optimizer type RMSProp and learning rate 0.0001, providing an accuracy average of up to 85.00%. Later, the CNN model can be used to establish a real-time, non-destructive, rapid, and precise purity detection system.

scholarly journals Classification of soybean tempe quality using deep learning

2021 ◽  
Vol 924 (1) ◽  
pp. 012022
Author(s):  
Y Hendrawan ◽  
B Rohmatulloh ◽  
I Prakoso ◽  
V Liana ◽  
M R Fauzy ◽  
...  
Keyword(s):  
Deep Learning ◽  
Classification Accuracy ◽  
Low Cost ◽  
Confusion Matrix ◽  
Learning Rate ◽  
Traditional Food ◽  
Quality Classification ◽  
Non Destructive

Abstract Tempe is a traditional food originating from Indonesia, which is made from the fermentation process of soybean using Rhizopus mold. The purpose of this study was to classify three quality levels of soybean tempe i.e., fresh, consumable, and non-consumable using a convolutional neural network (CNN) based deep learning. Four types of pre-trained networks CNN were used in this study i.e. SqueezeNet, GoogLeNet, ResNet50, and AlexNet. The sensitivity analysis showed the highest quality classification accuracy of soybean tempe was 100% can be achieved when using AlexNet with SGDm optimizer and learning rate of 0.0001; GoogLeNet with Adam optimizer and learning rate 0.0001, GoogLeNet with RMSProp optimizer, and learning rate 0.0001, ResNet50 with Adam optimizer and learning rate 0.00005, ResNet50 with Adam optimizer and learning rate 0.0001, and SqueezeNet with RSMProp optimizer and learning rate 0.0001. In further testing using testing-set data, the classification accuracy based on the confusion matrix reached 98.33%. The combination of the CNN model and the low-cost digital commercial camera can later be used to detect the quality of soybean tempe with the advantages of being non-destructive, rapid, accurate, low-cost, and real-time.

scholarly journals Classification of large green chilli maturity using deep learning

2021 ◽  
Vol 924 (1) ◽  
pp. 012009
Author(s):  
Y Hendrawan ◽  
B Rohmatulloh ◽  
I Prakoso ◽  
V Liana ◽  
M R Fauzy ◽  
...  
Keyword(s):  
Deep Learning ◽  
Small Molecules ◽  
Classification Accuracy ◽  
Low Cost ◽  
Confusion Matrix ◽  
Learning Rate ◽  
Maturity Index ◽  
Maturity Stage ◽  
Non Destructive

Abstract Chili (Capsicum annuum L.) is the source of various nutraceutical small molecules, such as ascorbic acid (vitamin C), carotenoids, tocopherols, flavonoids, and capsinoids. The purpose of this study was to classify the maturity stage of large green chili into three maturity levels, i.e. maturity 1 (maturity index 1 / 34 days after anthesis (DAA)), maturity 2 (maturity index 3 / 47 DAA), and maturity 3 (maturity index 5 / 60 DAA) by using convolutional neural networks (CNN) based deep learning and computer vision. Four types of pre-trained networks CNN were used in this study i.e. SqueezeNet, GoogLeNet, ResNet50, and AlexNet. From the overall sensitivity analysis results, the highest maturity classification accuracy of large green chili was 93.89% which can be achieved when using GoogLeNet with SGDmoptimizer and learning rate of 0.00005. However, in further testing using testing-set data, the highest classification accuracy based on confusion matrix was reaching 91.27% when using the CNN SqueezeNet model with RMSProp optimizer and a learning rate of 0.0001. The combination of the CNN model and the low-cost digital commercial camera can later be used to detect the maturity of large green chili with the advantages of being non-destructive, rapid, accurate, low-cost, and real-time.

scholarly journals An Automated Tomato Maturity Grading System Using Transfer Learning Based AlexNet

2021 ◽  
Vol 26 (2) ◽  
pp. 191-200
Author(s):  
Prasenjit Das ◽  
Jay Kant Pratap Singh Yadav ◽  
Arun Kumar Yadav
Keyword(s):  
Deep Learning ◽  
Transfer Learning ◽  
Low Cost ◽  
Confusion Matrix ◽  
Grading System ◽  
Learning Networks ◽  
Automatic Feature Extraction ◽  
Agriculture Sector ◽  
Extraction Behavior ◽  
Automated Grading

Tomato maturity classification is the process that classifies the tomatoes based on their maturity by its life cycle. It is green in color when it starts to grow; at its pre-ripening stage, it is Yellow, and when it is ripened, its color is Red. Thus, a tomato maturity classification task can be performed based on the color of tomatoes. Conventional skill-based methods cannot fulfill modern manufacturing management's precise selection criteria in the agriculture sector since they are time-consuming and have poor accuracy. The automatic feature extraction behavior of deep learning networks is most efficient in image classification and recognition tasks. Hence, this paper outlines an automated grading system for tomato maturity classification in terms of colors (Red, Green, Yellow) using the pre-trained network, namely 'AlexNet,' based on Transfer Learning. This study aims to formulate a low-cost solution with the best performance and accuracy for Tomato Maturity Grading. The results are gathered in terms of Accuracy, Loss curves, and confusion matrix. The results showed that the proposed model outperforms the other deep learning and the machine learning (ML) techniques used by researchers for tomato classification tasks in the last few years, obtaining 100% accuracy.

scholarly journals A Plastic Contamination Image Dataset for Deep Learning Model Development and Training

2020 ◽  
Vol 2 (2) ◽  
pp. 317-321
Author(s):  
Mathew G. Pelletier ◽  
Greg A. Holt ◽  
John D. Wanjura
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Detection System ◽  
Low Cost ◽  
Learning Algorithms ◽  
Machine Learning Algorithms ◽  
Image Dataset ◽  
The U.S ◽  
Image Datasets ◽  
Color Cameras

The removal of plastic contamination in cotton lint is an issue of top priority for the U.S. cotton industry. One of the main sources of plastic contamination appearing in marketable cotton bales is plastic used to wrap cotton modules on cotton harvesters. To help mitigate plastic contamination at the gin, automatic inspection systems are needed to detect and control removal systems. Due to significant cost constraints in the U.S. cotton ginning industry, the use of low-cost color cameras for detection of plastic contamination has been successfully adopted. However, some plastics of similar color to background are difficult to detect when utilizing traditional machine learning algorithms. Hence, current detection/removal system designs are not able to remove all plastics and there is still a need for better detection methods. Recent advances in deep learning convolutional neural networks (CNNs) show promise for enabling the use of low-cost color cameras for detection of objects of interest when placed against a background of similar color. They do this by mimicking the human visual detection system, focusing on differences in texture rather than color as the primary detection paradigm. The key to leveraging the CNNs is the development of extensive image datasets required for training. One of the impediments to this methodology is the need for large image datasets where each image must be annotated with bounding boxes that surround each object of interest. As this requirement is labor-intensive, there is significant value in these image datasets. This report details the included image dataset as well as the system design used to collect the images. For acquisition of the image dataset, a prototype detection system was developed and deployed into a commercial cotton gin where images were collected for the duration of the 2018–2019 ginning season. A discussion of the observational impact that the system had on reduction of plastic contamination at the commercial gin, utilizing traditional color-based machine learning algorithms, is also included.

scholarly journals A Zeroth-Order Adaptive Learning Rate Method to Reduce Cost of Hyperparameter Tuning for Deep Learning

2021 ◽  
Vol 11 (21) ◽  
pp. 10184
Author(s):  
Yanan Li ◽  
Xuebin Ren ◽  
Fangyuan Zhao ◽  
Shusen Yang
Keyword(s):  
Deep Learning ◽  
Network Architecture ◽  
Adaptive Methods ◽  
Learning Rate ◽  
Fine Tuning ◽  
Batch Size ◽  
Zeroth Order ◽  
Learning Rates ◽  
Initial Learning ◽  
Rate Method

Due to powerful data representation ability, deep learning has dramatically improved the state-of-the-art in many practical applications. However, the utility highly depends on fine-tuning of hyper-parameters, including learning rate, batch size, and network initialization. Although many first-order adaptive methods (e.g., Adam, Adagrad) have been proposed to adjust learning rate based on gradients, they are susceptible to the initial learning rate and network architecture. Therefore, the main challenge of using deep learning in practice is how to reduce the cost of tuning hyper-parameters. To address this, we propose a heuristic zeroth-order learning rate method, Adacomp, which adaptively adjusts the learning rate based only on values of the loss function. The main idea is that Adacomp penalizes large learning rates to ensure the convergence and compensates small learning rates to accelerate the training process. Therefore, Adacomp is robust to the initial learning rate. Extensive experiments, including comparison to six typically adaptive methods (Momentum, Adagrad, RMSprop, Adadelta, Adam, and Adamax) on several benchmark datasets for image classification tasks (MNIST, KMNIST, Fashion-MNIST, CIFAR-10, and CIFAR-100), were conducted. Experimental results show that Adacomp is not only robust to the initial learning rate but also to the network architecture, network initialization, and batch size.

Use of Deep Learning for Automatic Detection of Cracks in Tunnels: Prototype-2 Developed in the 2017–2018 Time Period

2019 ◽  
Vol 2673 (9) ◽  
pp. 44-50 ◽  
Author(s):  
Fred Daneshgaran ◽  
Luca Zacheo ◽  
Francesco Di Stasio ◽  
Marina Mondin
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Data Acquisition ◽  
Crack Detection ◽  
Detection System ◽  
Low Cost ◽  
Road Tunnel ◽  
Road Tunnels ◽  
Critical Problems ◽  
Academic Year

Cracks on the surfaces inside road tunnels are among the most critical problems in the management of such tunnels and, if not properly addressed, can have severe consequences in relation to safety and costs. Nowadays, the main techniques used for analysis of such surfaces make use of either human inspections or complex automated systems, which are, respectively, very time consuming and expensive, and/or difficult to implement. There is therefore great interest in a low-cost data-acquisition platform coupled with Artificial Intelligence-based automated crack detection system. This paper introduces a low-cost technique for road tunnel inspections based on a simple system that does not require complex preliminary work and can also be used in tunnels with a normal traffic flow. In particular, thanks to a second-generation data-acquisition system developed in the 2017–2018 academic year, a series of high-resolution pictures can be obtained and used in a pre-trained deep neural network able to identify the presence of cracks through the classification of the pictures. Thanks to deep learning techniques, it is possible to exploit the power of Inception-v4, a deep convolutional neural network provided by Google, which can be retrained for the specific purpose of crack detection. This kind of network has been trained with a pictures database generated using a second data-acquisition prototype developed during the 2017–2018 academic year.

scholarly journals Video Object Detection for Police Surveillance using Deep Learning

2021 ◽  
Vol 10 (3) ◽  
pp. 1-4
Keyword(s):  
Deep Learning ◽  
Urban Areas ◽  
Detection System ◽  
Confusion Matrix ◽  
Human Vision ◽  
Video Object ◽  
Television System ◽  
Public And Private ◽  
Human Operators ◽  
Develop System

Human vision is incredibly excellent and complex. In the previous years, people made significantly more leaps to expanding this visual capacity to machines. Cameras have been used as the eyes of computers.In response to increasing anxieties about crime and its threat to security and safety, the utilization of substantial numbers of closed-circuit television system (CCTV) in both public and private spaces have been considered a necessity. The use of these significant video footages is essential to incident investigations.But as the number of these systems rises, so as the need for human operator monitoring tasks.Unfortunately, many actionable incidents are utterly undetected in this manual systemdue to inherent limitations from deploying solely human operators eye-balling CCTV screens.As a result, surveillance footages are often used merely as passive records or as evidence for post-event investigations. This study aimed to develop a real-time firearm detection using deep learning embedded in CCTV cameras that pushes alert notifications to both iOS and Android mobile devices.This research used a descriptive design and asked IT experts to evaluate the develop system based on its compliance to ISO 25010 standard. Moreover, confusion matrix and intersection over union (IoU) were used to evaluate the performanceof the system.The detection system was found to be highly recommended in urban areas particularly for CCTVs found in barangay streets and establishments.

scholarly journals Application of Deep Learning in Integrated Pest Management: A Real-Time System for Detection and Diagnosis of Oilseed Rape Pests

2019 ◽  
Vol 2019 ◽  
pp. 1-14 ◽  
Author(s):  
Yong He ◽  
Hong Zeng ◽  
Yangyang Fan ◽  
Shuaisheng Ji ◽  
Jianjian Wu
Keyword(s):  
Deep Learning ◽  
Integrated Pest Management ◽  
Pest Management ◽  
Real Time ◽  
Oilseed Rape ◽  
Data Augmentation ◽  
Low Cost ◽  
Response Speed ◽  
Original Model ◽  
Real Time System

In this paper, we proposed an approach to detect oilseed rape pests based on deep learning, which improves the mean average precision (mAP) to 77.14%; the result increased by 9.7% with the original model. We adopt this model to mobile platform to let every farmer able to use this program, which will diagnose pests in real time and provide suggestions on pest controlling. We designed an oilseed rape pest imaging database with 12 typical oilseed rape pests and compared the performance of five models, SSD w/Inception is chosen as the optimal model. Moreover, for the purpose of the high mAP, we have used data augmentation (DA) and added a dropout layer. The experiments are performed on the Android application we developed, and the result shows that our approach surpasses the original model obviously and is helpful for integrated pest management. This application has improved environmental adaptability, response speed, and accuracy by contrast with the past works and has the advantage of low cost and simple operation, which are suitable for the pest monitoring mission of drones and Internet of Things (IoT).

Sign in / Sign up

Export Citation Format

Share Document