Matches in SemOpenAlex for { <https://semopenalex.org/work/W4367360280> ?p ?o ?g. }
Showing items 1 to 91 of
91
with 100 items per page.
- W4367360280 endingPage "102" @default.
- W4367360280 startingPage "89" @default.
- W4367360280 abstract "The amount of Diabetic Retinopathy (DR) patients is always increasing. There is a need to find a solution to prevent it, like regularly checking the eyes of patients to detect DR class. But the detection process also takes a long time and requires a large amount of money. So, this study proposed a deep learning method using CNN by implementing the MobilenetV2 architecture, which is a lightweight model and mobile-friendly architecture. In this study, we used the primary dataset from the APTOS 2019 dataset, which contains 3662 retinal fundus images. Diabetic retinopathy is divided into 5 classes, starting from No DR, Mild, Moderate, Severe, and Proliferative DR. To produce an optimal class classification in this study, first balance the dataset scheme to 800 in each class by using image augmentation and implementing circle cropping. So, the highest results are obtained with this method with the following parameters: batch size of 32, learning rate of 0.001, RMSProp optimizer, and epoch 100. Testing accuracy is 92.6% as a result of this parameter. The macro precision, recall, and f1-score are 92.8%, 92.6%, and 92.4%, respectively. Our results show that augmentation techniques and circle crop methods can make the MobileNetV2 architecture bring good performance results, and several parameters can bring significant system performance results. This research will provide information that MobileNetV2 has a good result on the parameters and this will be a lightweight model to classify DR into 5 classes." @default.
- W4367360280 created "2023-04-30" @default.
- W4367360280 creator A5004479698 @default.
- W4367360280 creator A5008411214 @default.
- W4367360280 creator A5032839399 @default.
- W4367360280 creator A5047129639 @default.
- W4367360280 creator A5048553247 @default.
- W4367360280 creator A5081424669 @default.
- W4367360280 date "2023-01-01" @default.
- W4367360280 modified "2023-09-26" @default.
- W4367360280 title "Diabetic Retinopathy Classification Based on Fundus Image Using Convolutional Neural Network (CNN) with MobilenetV2" @default.
- W4367360280 cites W2044517432 @default.
- W4367360280 cites W2114709213 @default.
- W4367360280 cites W2769713325 @default.
- W4367360280 cites W2805251526 @default.
- W4367360280 cites W2809489644 @default.
- W4367360280 cites W2911619450 @default.
- W4367360280 cites W2914960045 @default.
- W4367360280 cites W2963163009 @default.
- W4367360280 cites W2981771442 @default.
- W4367360280 cites W3033038270 @default.
- W4367360280 cites W3046531034 @default.
- W4367360280 cites W3117617661 @default.
- W4367360280 cites W3154979975 @default.
- W4367360280 cites W3156316671 @default.
- W4367360280 cites W3156453916 @default.
- W4367360280 cites W3162524694 @default.
- W4367360280 cites W3165183789 @default.
- W4367360280 cites W3186340003 @default.
- W4367360280 cites W3215768252 @default.
- W4367360280 cites W4200248162 @default.
- W4367360280 cites W4207003395 @default.
- W4367360280 cites W4214627663 @default.
- W4367360280 cites W4285104151 @default.
- W4367360280 doi "https://doi.org/10.1007/978-981-99-0248-4_7" @default.
- W4367360280 hasPublicationYear "2023" @default.
- W4367360280 type Work @default.
- W4367360280 citedByCount "0" @default.
- W4367360280 crossrefType "book-chapter" @default.
- W4367360280 hasAuthorship W4367360280A5004479698 @default.
- W4367360280 hasAuthorship W4367360280A5008411214 @default.
- W4367360280 hasAuthorship W4367360280A5032839399 @default.
- W4367360280 hasAuthorship W4367360280A5047129639 @default.
- W4367360280 hasAuthorship W4367360280A5048553247 @default.
- W4367360280 hasAuthorship W4367360280A5081424669 @default.
- W4367360280 hasConcept C108583219 @default.
- W4367360280 hasConcept C115961682 @default.
- W4367360280 hasConcept C118487528 @default.
- W4367360280 hasConcept C134018914 @default.
- W4367360280 hasConcept C153180895 @default.
- W4367360280 hasConcept C154945302 @default.
- W4367360280 hasConcept C2776391266 @default.
- W4367360280 hasConcept C2777212361 @default.
- W4367360280 hasConcept C2779829184 @default.
- W4367360280 hasConcept C31972630 @default.
- W4367360280 hasConcept C41008148 @default.
- W4367360280 hasConcept C555293320 @default.
- W4367360280 hasConcept C71924100 @default.
- W4367360280 hasConcept C81363708 @default.
- W4367360280 hasConceptScore W4367360280C108583219 @default.
- W4367360280 hasConceptScore W4367360280C115961682 @default.
- W4367360280 hasConceptScore W4367360280C118487528 @default.
- W4367360280 hasConceptScore W4367360280C134018914 @default.
- W4367360280 hasConceptScore W4367360280C153180895 @default.
- W4367360280 hasConceptScore W4367360280C154945302 @default.
- W4367360280 hasConceptScore W4367360280C2776391266 @default.
- W4367360280 hasConceptScore W4367360280C2777212361 @default.
- W4367360280 hasConceptScore W4367360280C2779829184 @default.
- W4367360280 hasConceptScore W4367360280C31972630 @default.
- W4367360280 hasConceptScore W4367360280C41008148 @default.
- W4367360280 hasConceptScore W4367360280C555293320 @default.
- W4367360280 hasConceptScore W4367360280C71924100 @default.
- W4367360280 hasConceptScore W4367360280C81363708 @default.
- W4367360280 hasLocation W43673602801 @default.
- W4367360280 hasOpenAccess W4367360280 @default.
- W4367360280 hasPrimaryLocation W43673602801 @default.
- W4367360280 hasRelatedWork W2297761292 @default.
- W4367360280 hasRelatedWork W2732542196 @default.
- W4367360280 hasRelatedWork W2738221750 @default.
- W4367360280 hasRelatedWork W2808969670 @default.
- W4367360280 hasRelatedWork W3120296212 @default.
- W4367360280 hasRelatedWork W3156786002 @default.
- W4367360280 hasRelatedWork W4293437635 @default.
- W4367360280 hasRelatedWork W4310880831 @default.
- W4367360280 hasRelatedWork W4324266846 @default.
- W4367360280 hasRelatedWork W564581980 @default.
- W4367360280 isParatext "false" @default.
- W4367360280 isRetracted "false" @default.
- W4367360280 workType "book-chapter" @default.