Online ISSN: 2515-8260

Keywords : CNN

Implementation on Privacy-Preserving Content-Based Image Retrieval in Cloud Image Repositories

Mr. Atish Gopichand Jadhav, Mr. Namdev Sawant, Mr. Subhash Pingale.

European Journal of Molecular & Clinical Medicine, 2023, Volume 10, Issue 1, Pages 3460-3471

Without knowing the name of the picture, searching through a collection of images that resemble the input images using a pursuing framework that uses the CBIR concept is essential. Overall, CBIR systems compare visual elements including colour, picture edge, surface, and the consistency of names between input images and images in the database. CNN is the characterisation method, while cosine comparability is used for recovery. This essay addresses the problem of large-scale image recovery, focusing on enhancing its accuracy and robustness. We focus on elements that might affect search vigour, such as different levels of illumination, object size and shape, fractional obstacles, and disordered foundations. These characteristics are particularly important when a hunt is conducted across extraordinarily huge datasets with high changeability. We suggest a brand-new CNN-based global descriptor termed REMAP, which is prepared from beginning to end with a triplet misfortune and learns and totals a progressive system of deep highlights from various CNN layers. REMAP categorically acquires discriminative cues that are typically constant and correlated at various semantic levels of visual reflection.


Dr. R. Kishore Kanna; U. Mutheeswaran; V. Subha Ramya; Dr. R. Vasuki; Dr. R Gomalavalli

European Journal of Molecular & Clinical Medicine, 2022, Volume 9, Issue 9, Pages 31-38

In medical practise, an electrocardiogram (ECG) is a crucial indicator tool for assessing cardiovascular arrhythmias. In this study, a machine learning system is used to compare patient ECGs and perform programmed ECG arrhythmia identification. The system was previously tuned based on an overall image informational index. Arrhythmias are more prevalent in those over the age of 60. A convolutional neural network (particularly, Alex Net) is utilised to extract features, and the highlights are then passed via a basic back spread neural network to finish the classification. The fundamental purpose of this research is to provide a simple, effective, and relevant learning strategy for categorising the three types of heart conditions (cardiac defects) so that a diagnosis may be made. The findings showed that when a moving deep learning highlight extractor was combined with a standard back proliferation neural architecture, very elite rates could be achieved. In a comparative analysis, validation accuracy was shown to be 100 percent in Google Net, 94 percent in Squeeze Net, and about 97.33 percent in Alex Net.


F. Ludyma Fernando, Dr. S. John Peter

European Journal of Molecular & Clinical Medicine, 2022, Volume 9, Issue 4, Pages 1960-1969

The affective quality called Valence refers to the intrinsic goodness (positive valence) or badness (negative valence) of an event, object, or situation. For this purpose, a model for classification and characterization of emotions have been developed which is discussed in this paper. In this model, the images are smoothened using an Average Filter and are first identified through a Convolutional Neural Network which uses the ReLU activation function. Then, the valence is classified using a Support Vector Machine (SVC) classifier, which uses a Radial Basis Function (RBF) kernel. For this reason, the emotions are labeled according to their nature. The positive emotions are labeled 1 (inclusive of the neutral emotion) and the negative emotions are labeled as 0. The images from the FER 2013 dataset is used for Valence Recognition and is given via a RBF Kernel in a SVM, which classifies whether the emotion recognized is positive or negative. The haarcascade algorithm is implemented to detect the face. In this paper, the 7 human emotions (happiness, surprise, fear, anger, fear, disgust, sadness and neutral) have been identified and their valence recognized.


J. Josphin Mary; R. Charanya; V. Shanthi; G. Sridevi

European Journal of Molecular & Clinical Medicine, 2020, Volume 7, Issue 9, Pages 1447-1453
DOI: 10.31838/ejmcm.07.09.154

Glaucoma is a persistent, permanent eye disease that contributes to vision and quality of life loss. Within this paper we build a deep learning system for the automatic diagnosis of glaucoma with a Convolutionary neural network. Deep learning algorithms, such as CNNs, that infer a hierarchical representation of images to differentiate between glaucoma and NG trends of diagnostic decisions. The DL architecture proposed contains six learning strategies: four Convolutionary strata and two entirely linked layers. Strategies for drop-out and data rise were implemented to further enhance the treatment of glaucoma. Extensive validation of ORIGA and SCES databases is carried out. The findings show that the recipient's operating curve field under curve (AUC) is significantly higher than the state of the art algorithms in glaucoma identification at 0,831 and 0,887 in the two databases. The method may be used for the detection of glaucoma.

Multispectral Image Classification Using Neural Networks in Astronomical Imagery

A. Praveena .; Nimisha Asthana; Anusha Chattopadhyay; Yash Singh

European Journal of Molecular & Clinical Medicine, 2020, Volume 7, Issue 2, Pages 5193-5206

multispectral image is an image that has wavelengths across the spectrum of electromagnetism. Astronomical images have various layers in their image capture process and thus fit into this category. This research aims to analyze the different lights of astronomical objects and their images and their colors. It focuses on how neural network models learn from each attribute of the images, thus aiming to find correlation and importance of the attributes of the images. Attempting to allow the CNN model to classify the images more efficiently by sorting the data.

Functionality of Pre-Prepared CNN Models using Deep Learning Technique for Detection of Parkinson Disease

Rohit Agarwal; Juginder Pal Singh

European Journal of Molecular & Clinical Medicine, 2020, Volume 7, Issue 4, Pages 121-125

Parkinson Disease is one of the most widely recognized neurodegenerative disorders. In
the U.S. Parkinson disease prevalence is roughly 20 cases for every 100,000 people per
year, with the mean period of beginning near 60 years. Thus, building up an automatic
system for detecting parkinson would be gainful for treating the infection without any
delay especially in remote areas. Due to the accomplishment of profound learning
calculations in breaking down clinical images, Convolution Neural Networks (CNNs)
have increased a lot of consideration for medical disease classification. What's more,
highlights realized by pre-prepared CNN models on huge scale datasets are a lot of
valuable in picture characterization errands. In this work, we evaluate the functionality of
pre-prepared CNN models used as highlight extractors followed by different classifiers
for the order of anomalous and normal MRI check pictures


Dr. Raja Sarath Kumar Boddu

European Journal of Molecular & Clinical Medicine, 2020, Volume 7, Issue 3, Pages 892-898

Brain tumour segmentation is one of the most critical and time-consuming jobs in the field of medical image processing since a human-assisted manual categorization may lead to incorrect prognosis and diagnosis. Furthermore, when there is a big quantity of data to be handled, it is a time-consuming job to say the least. There is a great deal of variation in brain tumours. There is a resemblance in appearance between tumour and normal tissues, which allows for the extraction of tumour areas from normal tissues. Images grow stubborn as time goes on. Using 2D Magnetic Resonance Imaging, we presented a technique for extracting brain tumours from brain scans. The Fuzzy C-Means clustering technique was used to cluster brain images (MRIs), which was then followed by conventional classifiers and other methods. A convolution neural network is a kind of neural network. The experimental research was conducted out on a real-time dataset including tumours of varying sizes, and Locations, forms, and varying picture intensities are all explored. In the conventional classifier section, we used six different traditional classifiers. Support Vector Machine (SVM), K-Nearest Neighbour (KNN), Multilayer Perception (MLP), and Logistic Regression are examples of machine learning algorithms. Regression, Nave Bayes, and Random Forest are all machine learning techniques that have been incorporated in scikit-learn. Following that, we went on to Convolution Neural Network (CNN) is a kind of neural network that is built using Keras and Tensor flow since it produces superior results. Performance as compared to the conventional ones CNN had an accuracy rate of 97.87 percent in our research, which is very impressive. The In this article, the primary objective is to differentiate between normal and aberrant pixels using texture-based and statistical methods. Characteristics that are based on

Parkinson's Disease Detection using Convolutional Neural Networks

Manisha Jindal; Yogesh Tripathi

European Journal of Molecular & Clinical Medicine, 2020, Volume 7, Issue 6, Pages 1298-1307

These days, a significant research exertion in social insurance biometrics is finding exact biomarkers that permit creating clinical choice help instruments. These instruments aid with the diagnosis and treatment of diseases such as Parkinson's disease. In this article, a convolutionary neural network (CNN) for the PD identification from drawing production is broken. This CNN comprises two parts: extraction and arranging (completely linked layers). CNN involves two pieces. CNN refers to the increase in frequency volume from 0 Hz to 25 Hz by the Fast Fourier Module. Throughout the modeling cycle the separating capacity of various headings tested achieved the greatest outcomes for both X & Y rollers. This research has been conducted using open database: a digital image tablet dataset from Parkinson Spiral Drawings. This study produced 96.5 percent of precision, 97.7 percent of F1 and 99.2 percent of region. There were the strongest results.

Detection and Identification of Forest Firing using Convolution Neural Network

Himanshu Sharma; Narendra Mohan

European Journal of Molecular & Clinical Medicine, 2020, Volume 7, Issue 4, Pages 126-129

Forest fire is a significant natural issue, making practical and biological harm while imperiling the human lives. The key component for controlling such marvel is fast identification .To accomplish this, one option is using neural networks to identify the fires, such that we implement Forest fire Detection. By using this convolution Neural Networks we detecting the fires that occur in the forest. Later we intimate message to forest officers then they take immediate action. CNN is a calculation that takes an input picture, assign the consequences (learnable loads and predispositions) to dissimilar perspectives in the picture and have the option to divide one from the other.

A Computer Aided Diagnosis of Lung Disease using Machine Learning Approach

Subapriya V; Jaichandran R; Shunmuganathan K.L; Abhiram Rajan; Akshay T; Shibil Rahman

European Journal of Molecular & Clinical Medicine, 2020, Volume 7, Issue 4, Pages 2662-2667

Cancer is a disease that is unregulated by cells in the body. Lung nodule is called lung cancer because the disease starts in the lungs. Cancer of the pulmonary system begins in the lungs and may travel to lymph nodes or other body species such as the brain. The lungs can also be impacted by cancer from other bodies. The metastases are named as cancer cells migrate from organ to organ. Lung cancers are normally grouped into two major cell and non-small cell types. In this study we predict a Computer Aided Diagnosis (CAD) for lung cancer prediction using Convolutional Neural Network (CNN) and ML approach

Helmet, Violation, Detection Using Deep Learning

Sherin Eliyas; K. Swaathi; Dr.P. Ranjana; A. Harshavardhan

European Journal of Molecular & Clinical Medicine, 2020, Volume 7, Issue 2, Pages 5173-5178

Road incidents are among the significant reasons, for the human passing. The majority of the passings in mishaps are because of harm to the top of the bike riders. Among the various sorts of street mishaps, bike mishaps are normal and cause extreme wounds. To reduce the involved risk for the motorcycle riders it is exceptionally fascinating to utilize helmet. The helmet is the motorcyclist's primary security. Many countries require the utilization of caps by motorcyclists, however numerous individuals neglect to comply with the law for different reasons. We present the advancement of a framework utilizing profound convolutional neural networks, (CNNs) for discovering bikers who are disregarding cap rules. The system involves motorcycle, detection, helmet, vs. no-helmet, classification, and method counting. Faster R-CNN with ResNet 50 network, model is implementing for motorcycle detector process. CNN classification model proposes for classify the helmet vs. no-helmet. Finally making alarm sound to alert the officer too preventing motorcycle accident. We assess the framework as far as accuracy and speed.