The Diagnosis Of Diabetic Retinopathy By Means Of Transfer Learning With Conventional Machine Learning Pipeline

Mekatronika ◽  
2020 ◽  
Vol 2 (2) ◽  
pp. 62-67
Author(s):  
Farhan Nabil Mohd Noor ◽  
Wan Hasbullah Mohd Isa ◽  
Anwar P.P. Abdul Majeed
Keyword(s):  
Machine Learning ◽  
Diabetic Retinopathy ◽  
Transfer Learning ◽  
Detection Method ◽  
Automated Detection ◽  
High Blood Glucose ◽  
Cotton Wool Spots ◽  
High Blood Glucose Level ◽  
Conventional Machine ◽  
The Common

Diabetic Retinopathy is one of the common eye diseases due to the complication of diabetes mellitus. Cotton wool spots, rough exudates, haemorrhages and microaneurysms are the symptoms of the diabetic retinopathy due to the fluid leakage that is caused by the high blood glucose level disorder. Early treatment to prevent a permanent blindness is important as it could save the diabetic retinopathy vision. Hence, in this study, we proposed to employ an automated detection method to diagnose the diabetic retinopathy. The dataset was obtained from the Kaggle Database and been divided for training, testing and validation purposes. Furthermore, Transfer Learning models, namely VGG19 were employed to extract the features before being processed by Machine Learning classifiers which are SVM, kNN and RF to classify the diabetic retinopathy. VGG19-SVM pipeline produced the best accuracy in training, testing and validation processes, achieving 99, 99 and 96 percents respectively.

Transfer learning based convolution neural net for authentication and classification of emotions from natural and stimulated speech signals

2021 ◽  
pp. 1-12
Author(s):  
Mukul Kumar ◽  
Nipun Katyal ◽  
Nersisson Ruban ◽  
Elena Lyakso ◽  
A. Mary Mekala ◽  
...  
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Transfer Learning ◽  
Research Work ◽  
Oral Communication ◽  
Neural Net ◽  
Emotion Classification ◽  
Style Transfer ◽  
Conventional Machine ◽  
Neural Network Algorithm

Over the years the need for differentiating various emotions from oral communication plays an important role in emotion based studies. There have been different algorithms to classify the kinds of emotion. Although there is no measure of fidelity of the emotion under consideration, which is primarily due to the reason that most of the readily available datasets that are annotated are produced by actors and not generated in real-world scenarios. Therefore, the predicted emotion lacks an important aspect called authenticity, which is whether an emotion is actual or stimulated. In this research work, we have developed a transfer learning and style transfer based hybrid convolutional neural network algorithm to classify the emotion as well as the fidelity of the emotion. The model is trained on features extracted from a dataset that contains stimulated as well as actual utterances. We have compared the developed algorithm with conventional machine learning and deep learning techniques by few metrics like accuracy, Precision, Recall and F1 score. The developed model performs much better than the conventional machine learning and deep learning models. The research aims to dive deeper into human emotion and make a model that understands it like humans do with precision, recall, F1 score values of 0.994, 0.996, 0.995 for speech authenticity and 0.992, 0.989, 0.99 for speech emotion classification respectively.

scholarly journals Diabetic Retinopathy Severity Detection using Convolutional Neural Network

2020 ◽  
Vol 32 ◽  
pp. 01012
Author(s):  
Mayank Shete ◽  
Saahil Sabnis ◽  
Srijan Rai ◽  
Gajanan Birajdar
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Diabetic Retinopathy ◽  
Convolutional Neural Network ◽  
Transfer Learning ◽  
Feature Learning ◽  
Eye Diseases ◽  
Accuracy Rate ◽  
Irreversible Damage ◽  
Computer Aided

Diabetic Retinopathy is one of the most prominent eye diseases and is the leading cause of blindness amongst adults. Automatic detection of Diabetic Retinopathy is important to prevent irreversible damage to the eye-sight. Existing feature learning methods have a lesser accuracy rate in computer aided diagnostics; this paper proposes a method to further increase the accuracy. Machine learning can be used effectively for the diagnosis of this disease. CNN and transfer learning are used for the severity classification and have achieved an accuracy of 73.9 percent. The use of XGBoost classifier yielded an accuracy of 76.5 percent.

scholarly journals Automated Detection of Multi-Rotor UAVs Using a Machine-Learning Approach

2020 ◽  
Vol 3 (3) ◽  
pp. 29
Author(s):  
Šimon Grác ◽  
Peter Beňo ◽  
František Duchoň ◽  
Martin Dekan ◽  
Michal Tölgyessy
Keyword(s):  
Machine Learning ◽  
Automated Detection ◽  
Learning Approach ◽  
Optimal Conditions ◽  
Detection Mechanism ◽  
Machine Learning Approach ◽  
Reliable Detection ◽  
The Common ◽  
Alternative Detection ◽  
Basic Experiments

The objective of this article is to propose and verify a reliable detection mechanism of multi-rotor unmanned aerial vehicles (UAVs). Such a task needs to be solved in many areas such as in the protection of vulnerable buildings or in the protection of privacy. Our system was firstly realized by standard computer vision methods using the Oriented FAST and Rotated BRIEF (ORB) feature detector. Due to the low success rate achieved in real-world conditions, the machine-learning approach was used as an alternative detection mechanism. The “Common Objects in Context dataset” was used as a predefined dataset and it was extended by 1000 samples of UAVs from the SafeShore dataset. The effectiveness and the reliability of our system are proven by four basic experiments—drone in a static image and videos which are displaying a drone in the sky, multiple drones in one image, and a drone with another flying object in the sky. The successful detection rate achieved was 97.3% in optimal conditions.

scholarly journals Classification of Watermelon Seeds Using Morphological Patterns of X-ray Imaging: A Comparison of Conventional Machine Learning and Deep Learning

Sensors ◽  
2020 ◽  
Vol 20 (23) ◽  
pp. 6753
Author(s):  
Mohammed Raju Ahmed ◽  
Jannat Yasmin ◽  
Eunsung Park ◽  
Geonwoo Kim ◽  
Moon S. Kim ◽  
...  
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Discriminant Analysis ◽  
Transfer Learning ◽  
Image Features ◽  
Linear Discriminant ◽  
X Ray ◽  
X Ray Imaging ◽  
Conventional Machine ◽  
Viable Seeds

In this study, conventional machine learning and deep leaning approaches were evaluated using X-ray imaging techniques for investigating the internal parameters (endosperm and air space) of three cultivars of watermelon seed. In the conventional machine learning, six types of image features were extracted after applying different types of image preprocessing, such as image intensity and contrast enhancement, and noise reduction. The sequential forward selection (SFS) method and Fisher objective function were used as the search strategy and feature optimization. Three classifiers were tested (linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), and k-nearest neighbors algorithm (KNN)) to find the best performer. On the other hand, in the transfer learning (deep learning) approaches, simple ConvNet, AlexNet, VGG-19, ResNet-50, and ResNet-101 were used to train the dataset and class prediction of the seed. For the supervised model development (both conventional machine learning and deep learning), the germination test results of the samples were used where the seeds were divided into two classes: (1) normal viable seeds and (2) nonviable and abnormal viable seeds. In the conventional classification, 83.6% accuracy was obtained by LDA using 48 features. ResNet-50 performed better than other transfer learning architectures, with an 87.3% accuracy which was the highest accuracy in all classification models. The findings of this study manifested that transfer learning is a constructive strategy for classifying seeds by analyzing their morphology, where X-ray imaging can be adopted as a potential imaging technique.

The Diagnosis of COVID-19 through X-ray Images via Transfer Learning and Fine-Tuned Dense Layer on Pipeline

Mekatronika ◽  
2021 ◽  
Vol 3 (2) ◽  
pp. 19-24
Author(s):  
Amiir Haamzah Mohamed Ismail ◽  
Mohd Azraai Mohd Razman ◽  
Ismail Mohd Khairuddin ◽  
Muhammad Amirul Abdullah ◽  
Rabiu Muazu Musa ◽  
...  
Keyword(s):  
Machine Learning ◽  
Transfer Learning ◽  
Screening Method ◽  
Learning Technologies ◽  
Fine Tuning ◽  
Dense Layer ◽  
X Ray ◽  
Chest X Ray ◽  
The Common ◽  
Fully Connected

X-ray is used in medical treatment as a method to diagnose the human body internally from diseases. Nevertheless, the development in machine learning technologies for pattern recognition have allowed machine learning of diagnosing diseases from chest X-ray images. One such diseases that are able to be detected by using X-ray is the COVID-19 coronavirus. This research investigates the diagnosis of COVID-19 through X-ray images by using transfer learning and fine-tuning of the fully connected layer. Next, hyperparameters such as dropout, p, number of neurons, and activation functions are investigated on which combinations of these hyperparameters will yield the highest classification accuracy model. InceptionV3 which is one of the common neural network is used for feature extraction from chest X-ray images. Subsequently, the loss and accuracy graphs are used to find the pipeline which performs the best in classification task. The findings in this research will open new possibilities in screening method for COVID-19.

scholarly journals Early Detection of Diabetic Retinopathy through Machine Learning Techniques

2019 ◽  
Vol 9 (2S) ◽  
pp. 444-447
Keyword(s):  
Machine Learning ◽  
Diabetic Retinopathy ◽  
Early Detection ◽  
Vital Role ◽  
Machine Learning Techniques ◽  
Fundus Image ◽  
Hard Exudates ◽  
Cotton Wool Spots ◽  
Chronic Heart Disease ◽  
Diabetic Eye Disease

Diabetic Retinopathy (DR) is progressive syndrome that leads to loss of vision if not detected and treated. Retina is inner tunic of the eyeball which is capillary and delicate transparent membrane. It is high developed tissue of eye which plays a major role for vision. Retina is the source for detection of many disorders. Part of retina with optic disc can be viewed through optamoloscope and termed as fundus image which is a basis of diagnosis for DR. DR can be categorized as Proliferative Diabetic Retinopathy (PDR), Diabetic Maculopathy, Nonproliferative Diabetic Retinopathy (NPDR) and Advanced Diabetic Eye Disease. Machine Learning (ML) techniques play a vital role in early detection of DR. In this paper a review on the existing techniques with open issues to be addressed is presented for diagnosing DR and model is proposed to consider the features namely Microaneurysms, Retinal Hemorrhages, Hard exudates, Cotton wool Spots, Neovascularization for classification of DR. These features can be combined with hypertension to predict other disorders like stroke, chronic heart disease, renal dysfunction, cardiovascular mortality and so on which overcome the need of other preliminary checkup .The complete profile of disorders for a diabetic patient can be deduced by the retinal fundus image.

scholarly journals Automated Detection and Differentiation of Drusen, Exudates, and Cotton-Wool Spots in Digital Color Fundus Photographs for Diabetic Retinopathy Diagnosis

2007 ◽  
Vol 48 (5) ◽  
pp. 2260 ◽  
Author(s):  
Meindert Niemeijer ◽  
Bram van Ginneken ◽  
Stephen R. Russell ◽  
Maria S. A. Suttorp-Schulten ◽  
Michael D. Abra`moff
Keyword(s):  
Diabetic Retinopathy ◽  
Automated Detection ◽  
Cotton Wool ◽  
Cotton Wool Spots ◽  
Fundus Photographs

scholarly journals Hybrid machine learning architecture for automated detection and grading of retinal images for diabetic retinopathy

2020 ◽  
Vol 7 (03) ◽  
pp. 1
Author(s):  
Barath Narayanan Narayanan ◽  
Russell C. Hardie ◽  
Manawaduge Supun De Silva ◽  
Nathaniel K. Kueterman
Keyword(s):  
Machine Learning ◽  
Diabetic Retinopathy ◽  
Automated Detection ◽  
Retinal Images ◽  
Hybrid Machine
Sign in / Sign up

Export Citation Format

Share Document