To see the other types of publications on this topic, follow the link: Deep learning CNN.
Journal articles on the topic 'Deep learning CNN'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 journal articles for your research on the topic 'Deep learning CNN.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse journal articles on a wide variety of disciplines and organise your bibliography correctly.
1
Aysuh, Jaggi, and Vinod Sharma Prof. "Classification of Healthy Seeds Using Deep Learning." Journal of Scientific Research and Technology (JSRT) 1, no. 4 (2023): 10–23. https://doi.org/10.5281/zenodo.8222793.
Full textAbstract:
With the increasing demand for healthy and high-quality seeds in agriculture, accurate and efficient seed classification methods are essential for seed quality control and optimisation of crop production. This work utilises a deep learning-based approach for healthy seed classification. It proposes a deep learning-based approach for beneficial seed classification, leveraging the power of neural networks to learn discriminative features from seed images automatically. The proposed method involves a multi-step pipeline that includes Image preprocessing, and Classification. The seed images are initially preprocessed to enhance their quality and reduce noise using image normalisation and denoising techniques. Next, a Deep convolutional neural network (CNN) is employed to extract relevant features from the preprocessed seed images. The CNN model is designed to capture the seeds' local and global characteristics, enabling it to learn complex patterns and textures.
2
Arora, Chinmay, Ritvik Gupta, and S. Sridhar. "Face Mask Detection using Deep Learning CNN Architecture." International Journal of Scientific Engineering and Research 10, no. 12 (2022): 1–10. https://doi.org/10.70729/se221206135738.
Full text
3
Dr., Rekha Patil, Kumar Katrabad Vidya, Mahantappa, and Kumar Sunil. "Image Classification Using CNN Model Based on Deep Learning." Journal Of Scientific Research And Technology (JSRT) 1, no. 2 (2023): 60–71. https://doi.org/10.5281/zenodo.7965526.
Full textAbstract:
In this work, we will use a convolutional neural network to classify images. In the field of visual image analysis, CNNs (a subset of deep neural networks) are the norm. Multilayer perceptron is used to develop CNN; it is based on a hierarchical model that works on network construction and then delivers to a fully linked layer. All the neurons are linked together and their output is processed in this layer. Here, we demonstrate how our system can get the job done in challenging domains like computer vision by using a deep learning approach. Convolutional Neural Networks (CNNs) are a machine learning method employed by our system for automated picture categorization. For grayscale picture categorization, our method compares to the Digit of MNIST data set. More processing power is needed for picture classification because of the grayscale images in the training data set. Our model's great accuracy in picture classification can be seen in the experimental phase, where we trained it using a convolutional neural network and obtained a result of 98% accuracy
4
Santosh, Giri1 and Basanta Joshi. "TRANSFER LEARNING BASED IMAGE VISUALIZATION USING CNN." International Journal of Artificial Intelligence and Applications (IJAIA) 10, July (2019): 47–55. https://doi.org/10.5281/zenodo.3371299.
Full textAbstract:
Image classification is a popular machine learning based applications of deep learning. Deep learning techniques are very popular because they can be effectively used in performing operations on image data in large-scale. In this paper CNN model was designed to better classify images. We make use of featureextraction part of inception v3 model for feature vector calculation and retrained the classification layer with these feature vector. By using the transfer learning mechanism the classification layer of the CNN model was trained with 20 classes of Caltech101 image dataset and 17 classes of Oxford 17 flower image dataset. After training, network was evaluated with testing dataset images from Oxford 17 flower dataset and Caltech101 image dataset. The mean testing precision of the neural network architecture with Caltech101 dataset was 98 % and with Oxford 17 Flower image dataset was 92.27 %.
5
Mohebbanaaz, Mohebbanaaz, Y. Padma Sai, and L. V. Rajani Kumari. "Detection of cardiac arrhythmia using deep CNN and optimized SVM." Indonesian Journal of Electrical Engineering and Computer Science 24, no. 1 (2021): 217–25. https://doi.org/10.11591/ijeecs.v24.i1.pp217-225.
Full textAbstract:
Deep learning (DL) has become a topic of study in various applications, including healthcare. Detection of abnormalities in an electrocardiogram (ECG) plays a significant role in patient monitoring. It is noted that a deep neural network when trained on huge data, can easily detect cardiac arrhythmia. This may help cardiologists to start treatment as early as possible. This paper proposes a new deep learning model adapting the concept of transfer learning to extract deep-CNN features and facilitates automated classification of electrocardiogram (ECG) into sixteen types of ECG beats using an optimized support vector machine (SVM). The proposed strategy begins with gathering ECG datasets, removal of noise from ECG signals, and extracting beats from denoised ECG signals. Feature extraction is done using ResNet18 via concept of transfer learning. These extracted features are classified using optimized SVM. These methods are evaluated and tested on the MIT-BIH arrhythmia database. Our proposed model is effective compared to all State of Art Techniques with an accuracy of 98.70%.
6
Ahmed, M. Alkababji, and H. Mohammed Omar. "Real time ear recognition using deep learning." TELKOMNIKA Telecommunication, Computing, Electronics and Control 19, no. 2 (2021): pp. 523~530. https://doi.org/10.12928/TELKOMNIKA.v19i2.18322.
Full textAbstract:
Automatic identity recognition of ear images represents an active area of interest within the biometric community. The human ear is a perfect source of data for passive person identification. Ear images can be captured from a distance and in a covert manner; this makes ear recognition technology an attractive choice for security applications and surveillance in addition to related application domains. Differing from other biometric modalities, the human ear is neither affected by expressions like faces are nor do need closer touching like fingerprints do. In this paper, a deep learning object detector called faster region based convolutional neural networks (Faster R-CNN) is used for ear detection. A convolutional neural network (CNN) is used as feature extraction. principal component analysis (PCA) and genetic algorithm are used for feature reduction and selection respectively and a fully connected artificial neural network as a matcher. The testing proved the accuracy of 97.8% percentage of success with acceptable speed and it confirmed the accuracy and robustness of the proposed system.
7
Prof., A. R. Ghongade Sneha Zade Yash Malankar Sameer Kamble Pranali Dhenge. "Object Caption Generator Using Deep Learning." International Journal of Advanced Innovative Technology in Engineering 9, no. 3 (2024): 324–28. https://doi.org/10.5281/zenodo.12747531.
Full textAbstract:
In this project, we use CNN and LSTM to identify the caption of the object. As the deep learning techniques are growing, huge datasets and computer power are helpful to build models that can generate captions for an object. This is what we are going to implement in this Python based project where we will use deep learning techniques like CNN and RNN. Object caption generator is a process which involves natural language processing and computer vision concepts to recognize the context of an object and present it in English. In this survey, we carefully follow some of the core concepts of object captioning and its common approaches. We discuss Keras library, numpy and Pycharm for the making of this project.We also discuss about flickr_dataset and CNN used for object classification. The system is trained on a large dataset of objects and their corresponding captions, using techniques such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs). The CNNs are used to extract features from the objects, while the RNNs are used to generate the textual descriptions. The object caption generator is a promising application of machine learning in the field of computer vision. It has many potential uses, including assisting the visually impaired, creating better search results for object-based queries, and helping with content creation for social media and marketing purposes.
8
A., Sasi Kumar, and S. Aithal P. "DeepQ Residue Analysis of Brain-Computer Classification and Prediction using Deep CNN." International Journal of Applied Engineering and Management Letters (IJAEML) 7, no. 2 (2023): 144–63. https://doi.org/10.5281/zenodo.8104434.
Full textAbstract:
<strong>Purpose: </strong><em>During this</em><em> article, we are going to consistently explore the kinds of brain signals for Brain Computer Interface (BCI) and discover the related ideas of the in-depth learning of brain signal analysis. We talk review recent machine Associate in Nursing deep learning approaches within the detection of two brain unwellness just like Alzheimer' disease (AD), brain tumor. In addition, a quick outline of the varied marker extraction techniques that want to characterize brain diseases is provided. Project work, the automated tool for tumor classification supported by image resonance information. It is given by various convolutional neural network (CNN) samples with ResNet Squeeze.</em> <strong>Objectives: </strong><em>This paper is to analyse brain diseases classification and prediction using deep learning concepts</em>. <em>Deep learning is a group of machine learning in computer science that has networks capable of unattended learning from data that's unstructured or unlabelled. conjointly called deep neural learning could be a operation of Al that mimics however, the human brain works in process data to be used in object detection, speech recognition, language translation, and call making. </em> <strong>Methodology: </strong><em>To test the result by measuring the semantics in the input sentence, the creation of embedded vectors with the same value is achieved. In this case, a sentence with a different meaning is used. Since it is difficult to collect a large amount of labelled data, it simulates the signal in different sentences. As you progress, teach for extra complicated capabilities with layers from the shared output of preceding layers. We examine forms of deep getting to know methods: LSTM Model with RNN, CNN results. CNN is a multi-layer feed-ahead neural community. The gadget weight is up to date via way of means of the Backpropagation Error procedure. TF-IDF of time period t in record d. Unlike traditional precis models, the ahead engineering feature is predicated on understanding of the required records area. In addition, this framework is related to synthetic abbreviations, which might be then used to put off the impact of guide function improvement and records labelling.</em> <strong>Results: </strong><em>We will follow this option of 257 factors as vector enter category algorithms. It is a aggregate of the subsequent forms with enter layer, convolution layer, linear unit (ReLU) layer, pooling layer, absolutely coupled layer. A recurrent neural community (RNN) is a form of a neural community that defines connections among loop units. This creates an inner community country that allows. Feature choice is a extensively used approach that improves the overall performance of classifiers. Here, we examine the consequences of conventional magnificence fires with correlation-primarily based totally man or woman choice. </em> <strong>Originality: </strong><em>Analysis of Brain Diseases with the approach of Computer Classification and Prediction using Deep CNN </em><em>with ResNet Squeeze.</em> <strong>Type of Paper:</strong> <em>Conceptual research paper.</em>
9
Pravallika, V., V. Uday Kiran, B. Rahul, N. Neelima, G. Rishi Patnaik, and DR Sreejyothshna Ankam. "Deep Learning-Based Image Captioning: A Hybrid CNN-LSTM Approach." International Journal of Research Publication and Reviews 6, no. 4 (2025): 2459–63. https://doi.org/10.55248/gengpi.6.0425.1392.
Full text
10
Gupta, Jaya, Sunil Pathak, and Gireesh Kumar. "Deep Learning (CNN) and Transfer Learning: A Review." Journal of Physics: Conference Series 2273, no. 1 (2022): 012029. http://dx.doi.org/10.1088/1742-6596/2273/1/012029.
Full textAbstract:
Abstract Deep Learning is a machine learning area that has recently been used in a variety of industries. Unsupervised, semi-supervised, and supervised-learning are only a few of the strategies that have been developed to accommodate different types of learning. A number of experiments showed that deep learning systems fared better than traditional ones when it came to image processing, computer vision, and pattern recognition. Several real-world applications and hierarchical systems have utilised transfer learning and deep learning algorithms for pattern recognition and classification tasks. Real-world machine learning settings, on the other hand, often do not support this assumption since training data can be difficult or expensive to get, and there is a constant need to generate high-performance beginners who can work with data from a variety of sources. The objective of this paper is using deep learning to uncover higher-level representational features, to clearly explain transfer learning, to provide current solutions and evaluate applications in diverse areas of transfer learning as well as deep learning.
11
Gupta, Jaya, Sunil Pathak, and Gireesh Kumar. "Deep Learning (CNN) and Transfer Learning: A Review." Journal of Physics: Conference Series 2273, no. 1 (2022): 012029. http://dx.doi.org/10.1088/1742-6596/2273/1/012029.
Full textAbstract:
Abstract Deep Learning is a machine learning area that has recently been used in a variety of industries. Unsupervised, semi-supervised, and supervised-learning are only a few of the strategies that have been developed to accommodate different types of learning. A number of experiments showed that deep learning systems fared better than traditional ones when it came to image processing, computer vision, and pattern recognition. Several real-world applications and hierarchical systems have utilised transfer learning and deep learning algorithms for pattern recognition and classification tasks. Real-world machine learning settings, on the other hand, often do not support this assumption since training data can be difficult or expensive to get, and there is a constant need to generate high-performance beginners who can work with data from a variety of sources. The objective of this paper is using deep learning to uncover higher-level representational features, to clearly explain transfer learning, to provide current solutions and evaluate applications in diverse areas of transfer learning as well as deep learning.
12
Muhammad, Zulqarnain, Ghazali Rozaida, Mazwin Mohmad Hassim Yana, and Rehan Muhammad. "A comparative review on deep learning models for text classification." Indonesian Journal of Electrical Engineering and Computer Science (IJEECS) 19, no. 1 (2020): 325–35. https://doi.org/10.11591/ijeecs.v19.i1.pp325-335.
Full textAbstract:
Text classification is a fundamental task in several areas of natural language processing (NLP), including words semantic classification, sentiment analysis, question answering, or dialog management. This paper investigates three basic architectures of deep learning models for the tasks of text classification: Deep Belief Neural (DBN), “Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN), these three main types of deep learning architectures, are largely explored to handle various classification tasks. DBN have excellent learning capabilities to extracts highly distinguishable features and good for general purpose. CNN have supposed to be better at extracting the position of various related features while RNN is modeling in sequential of long-term dependencies. This paper work shows the systematic comparison of DBN, CNN, and RNN on text classification tasks. Finally, we show the results of deep models by research experiment. The aim of this paper to provide basic guidance about the deep learning models that which models is best for the task of text classification.
13
Sriwong, Kittipat, Kittisak Kerdprasop, and Nittaya Kerdprasop. "The Study of Noise Effect on CNN-Based Deep Learning from Medical Images." International Journal of Machine Learning and Computing 11, no. 3 (2021): 202–7. http://dx.doi.org/10.18178/ijmlc.2021.11.3.1036.
Full textAbstract:
Currently, computational modeling methods based on machine learning techniques in medical imaging are gaining more and more interests from health science researchers and practitioners. The high interest is due to efficiency of modern algorithms such as convolutional neural networks (CNN) and other types of deep learning. CNN is the most popular deep learning algorithm because of its prominent capability on learning key features from images that help capturing the correct class of images. Moreover, several sophisticated CNN architectures with many learning layers are available in the cloud computing environment. In this study, we are interested in performing empirical research work to compare performance of CNNs when they are dealing with noisy medical images. We design a comparative study to observe performance of the AlexNet CNN model on classifying diseases from medical images of two types: images with noise and images without noise. For the case of noisy images, the data had been further separated into two groups: a group of images that noises harmoniously cover the area of the disease symptoms (NIH) and a group of images that noises do not harmoniously cover the area of the disease symptoms (NNIH). The experimental results reveal that NNIH has insignificant effect toward the performance of CNN. For the group of NIH, we notice some effect of noise on CNN learning performance. In NIH group of images, the data preparation process before learning can improve the efficiency of CNN.
14
Bindushree, S., and A. N. Rakshitha. "Face Recognition Using Deep Learning." International Journal of Advanced Scientific Inovation 01, no. 01 (2021): 12–18. https://doi.org/10.5281/zenodo.4641691.
Full textAbstract:
<strong><em>Today face recognition and its usage are developing at a remarkable rate. Researches are at present building up different strategies in which facial recognition framework works. In circumstances like accidents, normal disasters, missing cases, clashes between nations, kidnappings and numerous different circumstances individuals are regularly isolated by their families. Recognizing the relatives of those refugees is essential to arrive at their family for refugee’s security and backing. Everyday polices are enrolling with missing cases, a portion of those enlisted cases are getting tackled and some are definitely not by using the manual method where it takes more time. The goal of this paper is to provide a solution to overcome time delay from existing strategies for police examination utilizing most recent innovation. Hence we adopt a framework which utilizes CNN (Convolutional Neural Network) technique with VGG16 architecture where we use our raw dataset which contains 84 images collected from 21 families data, after applying augmentation method the image count in final dataset is increased to 1512, then from this dataset 80% of data is used for training data and 20% is used for testing data. This framework helps to verify an individual’s trait using their face and family subtleties with related model with increased accuracy, and gives a effective solution for identifying refugee’s family.</em></strong>
15
S.Mahalakshmi and Dheeba J. Dr. "Robust Approach of Automatic Number Plate Recognition System using Deep CNN." JXU Journal, Scopus Indexed 50, no. 3 (2023): 1–5. https://doi.org/10.5281/zenodo.8382498.
Full textAbstract:
<strong>Automatic License plate / Number plate / Registration plate recognition is recognized as an automation which evolved mostly based on image processing techniques. It has been extensively used in recognizing vehicles in applications such as red-light enforcement, over speeding, parking control, toll collection. The main objective of the paper is to identify the most well-planned way to identify the registration plate from the digital image (gained from the camera)and recognize with high accuracy. ANPR is employed to localize the license plates, segment each character and extract the text from the license plate and then recognize each character successfully. The main issue of registration plate recognition relies on the accuracy rate. Advancement in deep learning methods has improved the ability to solve visual recognition tasks. Henceforth using deep Convolutional Neural Networks (DCNN) will intensify the precision, recall, processing speed, and reduce the error rate in solving the ANPR process. The use of deep learning CNN helps in identification of license plates of any aspect ratio which would work well for places like India where license plate style differs remarkably. The CNNs are highly skilled and balanced so that they are strong under various states like variations in pose, lighting, occlusion etc. In our dataset we have used 100 images to train our network and obtained 99% accuracy for plate localization and 93% accuracy for recognition.</strong>
16
Neetha, Alex, Lal Abhijith, S. Saranya, Anna Sunil Sharon, and S. Sreejith. "Prediction of Chronic Disease using Deep Learning." Recent Trends in Androids and IOS Applications 6, no. 1 (2023): 16–22. https://doi.org/10.5281/zenodo.10205658.
Full textAbstract:
<i>In present times, people are exposed to various illnesses due to their lifestyle and the state of the environment. It is crucial to identify and predict these diseasesat an early stage to prevent their progression to a more severe state. The goal of this proposed work is to identify and predict the patientswith more common chronicillness. The prediction gives thebenefit of early disease detection. In this proposed work, prediction is done by deep learning. The paper proposesa deep learning approach for predicting chronicdiseases using a convolution neural network (CNN) model. The system comprises three modules: an admin module, a doctor module and a patient module. The admin module provides an interface for managing patient data, while the doctor module allows doctor to access patientdata and generatereports. The patientmodule allows the patient to input their health data and receivepersonalized health recommendations. The CNN model is used to learn complex pattern in the patient data and predict the risk of chronic disease.The proposed system has the potential to improve the accuracy of chronic disease prediction, enabling earlyintervention and prevention of these diseases.</i>
17
Zahraa, Najm Abdullah, Abdulridha Abutiheen Zinah, A. Abdulmunem Ashwan, and A. Harjan Zahraa. "Official logo recognition based on multilayer convolutional neural network model." TELKOMNIKA (Telecommunication, Computing, Electronics and Control) 20, no. 5 (2022): 1083–90. https://doi.org/10.12928/telkomnika.v20i5.23464.
Full textAbstract:
Deep learning has gained high popularity in the field of image processing and computer vision applications due to its unique feature extraction property. For this characteristic, deep learning networks used to solve different issues in computer vision applications. In this paper the issue has been raised is classification of logo of formal directors in Iraqi government. The paper proposes a multi-layer convolutional neural network (CNN) to classify and recognize these official logos by train the CNN model on several logos. The experimental show the effectiveness of the proposed method to recognize the logo with high accuracy rate about 99.16%. The proposed multi-layers CNN model proves the effectiveness to classify different logos with various conditions.
18
Pavithra, G.K., and S. Shridevi. "Extraction of Ship Images using Deep Learning." International Journal of Engineering and Advanced Technology (IJEAT) 9, no. 5 (2020): 635–38. https://doi.org/10.35940/ijeat.E9682.069520.
Full textAbstract:
Ship Extraction is very important in the marine industry. Extraction of ships is helpful to the fishers to find the other ships nearly around the particular area. Still today the fishers are to find the ships using some traditional methods. But now it became difficult due to environmental changes. So, by using the deep learning techniques like the CNN algorithm the ship extraction can be identified effectively. Generally, the ships are identified as narrow bow and parallel hull edge, etc. Here, the Existing system they have used the Tensor flow, to see the performance of the datasets, using Recall and precision. In the proposed system, we are using CNN deep learning techniques to identify the ships. By finding the ships with the techniques, the time will be saved and the productivity can be increased. The features of the ship image are taken and trained using the neural network algorithm and then the prediction is done by testing the images.
19
K, Manimekalai, and A. Kavitha Dr. "Deep Learning Methods in Classification of Myocardial Infarction by employing ECG Signals." Indian Journal of Science and Technology 13, no. 28 (2020): 2823–32. https://doi.org/10.17485/IJST/v13i28.445.
Full textAbstract:
Abstract <strong>Background/Objectives:</strong> To automatically classify and detect the Myocardial Infarction using ECG signals.<strong> Methods/Statistical analysis:</strong> Deep Learning algorithms Convolutional Neural Network(CNN), Long Short Term Memory(LSTM) and Enhanced Deep Neural Network(EDN) were implemented. The proposed model EDN, comprises the techniques CNN and LSTM. Vector operations like matrix multiplication and gradient decent were applied to large matrices of data that are executed in parallel with GPU support. Because of parallelism EDN faster the execution time of process. <strong>Findings:</strong> Proposed model EDN yields better accuracy (88.89%) than other state-of-art methods for PTB database. <strong>Novelty/Applications:</strong> The proposed classification algorithm for analyzing the ECG signals is obtained by comprising the Convolutional Neural Network(CNN)and Long short-term memory networks(LSTM). Also, it is identified that the novel classification technique based on deep learning decreases the misdiagnosis rate of MI. <strong>Keywords:</strong> Classification; CNN; deep learning; deep neural network; EDN; LSTM; Myocardial Infarction(MI)
20
Binjal, Suthar, and Gadhiya Bijal. "Child Activity Recognition using Deep Learning." International Journal of Engineering and Advanced Technology (IJEAT) 9, no. 5 (2020): 364–67. https://doi.org/10.35940/ijeat.E9563.069520.
Full textAbstract:
The human action recognition is the subject to predicting what an individual is performing based on a trace of their development exploiting a several strategies. Perceiving human activities is an ordinary region of eagerness in view of its various potential applications; though, it is still in start. It is a trending analysis area possessed by the range from dependable automation, medicinal services to developing the smart supervision system. In this work, we are trying to recognize the activity of the child from video dataset using deep learning techniques. The proposed system will help parent to take care of their baby during the job or from anywhere else to know what the baby is doing. This can also be useful to prevent the in-house accident falls of the child and for health monitoring. The activities can be performed by child include sleeping, walking, running, crawling, playing, eating, cruising, clapping, laughing, crying and many more. We are focusing on recognizing crawling, running, sleeping, and walking activities of the child in this study. The offered system gives the best result compared with the existing methods, which utilize sensor-based information. Experimental results proved that the offered deep learning model had accomplished 94.73% accuracy for recognizing the child activity.
21
Asadullah, Kehar, Hussain Arain Rafaqat, and Ahmed Shaikh Riaz. "Deciphering complex text-based CAPTCHAs with deep learning." Indian Journal of Science and Technology 13, no. 13 (2020): 1390–400. https://doi.org/10.17485/IJST/v13i13.126.
Full textAbstract:
Abstract <strong>Background:</strong> CAPTCHA is a mechanism to distinguish humans from bots. It has become standard means of protection from the misuse of resources on World Wide Web. Different types of CAPTCHAs are implemented but text-based schemes are the most widely used due to its easiness and robustness. A user is asked to type in the text from an image. The image is intentionally distorted to dodge the bots. Recognizing the text is easy for humans but very hard for computers. <strong>Method/Findings:</strong> In this work, a text-based CAPTCHA scheme with background clutter and partially connected characters is decoded. The main steps consist on preprocessing, segmentation and recognition. Several digital image processing techniques were applied during preprocessing, segmentation steps and convolutional neural network (CNN) was used for recognition process. Since massive data is required for CNN therefore data was generated synthetically. A complex text-based CAPTCHA scheme with varying number of letters: 3, 4 and 5 letters is decoded with the overall precision of 77.5%, 64.2% and 51.9% respectively. <strong>Keywords:</strong> CAPTCHAs; HIPs; image processing; machine learning; CNN
22
Srinivas, Dr Kalyanapu, and Reddy Dr.B.R.S. "Deep Learning based CNN Optimization Model for MR Braing Image Segmentation." Journal of Advanced Research in Dynamical and Control Systems 11, no. 11 (2019): 213–20. http://dx.doi.org/10.5373/jardcs/v11i11/20193190.
Full text
23
Muhaafidz, Md Saufi, Afiq Zamanhuri Mohd, Mohammad Norasiah, and Ibrahim Zaidah. "Deep Learning for Roman Handwritten Character Recognition." Indonesian Journal of Electrical Engineering and Computer Science 12, no. 2 (2018): 455–60. https://doi.org/10.11591/ijeecs.v12.i2.pp455-460.
Full textAbstract:
The advantage of deep learning is that the analysis and learning of massive amounts of unsupervised data make it a beneficial tool for Big Data analysis. Convolution Neural Network (CNN) is a deep learning method that can be used to classify image, cluster them by similarity, and perform image recognition in the scene. This paper conducts a comparative study between three deep learning models, which are simple-CNN, AlexNet and GoogLeNet for Roman handwritten character recognition using Chars74K dataset. The produced results indicate that GooleNet achieves the best accuracy but it requires a longer time to achieve such result while AlexNet produces less accurate result but at a faster rate.
24
Et. al., Mihir Verma,. "Action Recognition Using Deep Learning And Cnn." Turkish Journal of Computer and Mathematics Education (TURCOMAT) 12, no. 11 (2021): 818–24. http://dx.doi.org/10.17762/turcomat.v12i11.5967.
Full textAbstract:
Automated action recognition using Deep learning and CNN is playing a vital role in today‘s day to day society, it may be video action recognitions through cctv, or it may be the smart homes. Now day’s human actions are used in many devices to control them like HoloLens VR, for that recognition of action is important that why video recognition. This Paper represents practical, reliable, and generic systems for video-based human action recognition, technology of CNN network is used to recognize different layers of the video images features. These features are obtained by extracting the features from different layers that are through the CNN (Convolutional Neural Network).
25
Anbarasi., A., and S. Ravi. "Optimal Deep Learning based Classification Model for Mitral Valve Diagnosis System." International Journal of Engineering and Advanced Technology (IJEAT) 9, no. 4 (2020): 315–21. https://doi.org/10.35940/ijeat.C6530.049420.
Full textAbstract:
In present days, the domain of mitral valve (MV) diagnosis so common due to the changing lifestyle in day to day life. The increased number of MV disease necessitates the development of automated disease diagnosis model based on segmentation and classification. This paper makes use of deep learning (DL) model to develop a MV classification model to diagnose the severity level. For the accurate classification of ML, this paper applies the DL model called convolution neural network (CNN-MV) model. And, an edge detection based segmentation model is also applied which will helps to further enhance the performance of the classifier. Due to the non-availability of MV dataset, we have collected a MV dataset of our own from a total of 211 instances. A set of three validation parameters namely accuracy, sensitivity and specificity are applied to indicate the effective operation of the CNN-MV model. The obtained simulation outcome pointed out that the presented CNN-MV model functions as an appropriate tool for MV diagnosis.
26
Zhan, Zhiwei, Guoliang Liao, Xiang Ren, et al. "RA-CNN." International Journal of Software Science and Computational Intelligence 14, no. 1 (2022): 1–14. http://dx.doi.org/10.4018/ijssci.311446.
Full textAbstract:
Emotion is a feeling that can be expressed by different mediums. Emotion analysis is a key task in NLP which is responsible for judging the emotional tendency of texts. Currently, in a complex multi-semantic environment, it still suffers from poor performance. Traditional methods usually require human intervention, while deep learning always has a trade-off between local and global features. To solve the problem that deep learning models generalize poorly for emotion analysis, this article proposed a semantic-enhanced method called RA-CNN, a classification model under a multi-semantic environment. It integrates CNN for local feature extraction, RNN for global feature extraction, and attention mechanism for feature scaling. As a result, it can acquire the correct meaning of sentences. After experimenting with the hotel review dataset, it has an improvement in positive feeling classification compared with the baseline model (3%~13%), and it showed a competitive performance compared with ordinary deep learning models (~1%). On negative feeling classification, it also performed well close to other models.
27
Park, Sungwoo, Kyong-Ho Han, and Wooyoung Jang. "CNN Deep Learning Acceleration Algorithm for Mobile System." Journal of Korean Institute of Information Technology 16, no. 10 (2018): 1–9. http://dx.doi.org/10.14801/jkiit.2018.16.10.1.
Full text
28
Sonata, I., Y. Heryadi, L. Lukas, and A. Wibowo. "Autonomous car using CNN deep learning algorithm." Journal of Physics: Conference Series 1869, no. 1 (2021): 012071. http://dx.doi.org/10.1088/1742-6596/1869/1/012071.
Full text
29
Venkata, Gogireddy, Nerella Ganesh, Potturi Teja, and Senthil Kumar. "Object Detection through CNN with Deep Learning." International Journal of Computer Applications 176, no. 15 (2020): 46–49. http://dx.doi.org/10.5120/ijca2020920126.
Full text
30
Sumahasan, S. "Object Detection using Deep Learning Algorithm CNN." International Journal for Research in Applied Science and Engineering Technology 8, no. 7 (2020): 1578–84. http://dx.doi.org/10.22214/ijraset.2020.30594.
Full text
31
Shaikh, Mohd Salim, Lucky Nirankari, Vasant Pardeshi, Rupesh Sharma, and Prof Sunil Kale. "DEEPFAKE DETECTION USING DEEP LEARNING (CNN+LSTM)." INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT 07, no. 11 (2023): 1–11. http://dx.doi.org/10.55041/ijsrem26808.
Full textAbstract:
Artificial intelligence advancements have led to the development of deepfake technology, which seriously jeopardises the integrity of visual media material. Robust detection algorithms are becoming more and more necessary as deepfake creation techniques become more complex. This study combines Long Short-Term Memory (LSTM) networks with Convolutional Neural Networks (CNNs) to present a novel method for deepfake identification. The suggested CNN+LSTM architecture makes use of LSTMs' temporal modelling capabilities and CNNs' spatial feature extraction capabilities. While the LSTM component analyses the temporal connections between frames to identify patterns suggestive of deepfake manipulation, the CNN component concentrates on capturing local features and patterns in individual frames. The combination of these two networks improves the model's capacity to identify minute anomalies and inconsistencies that are indicative of deepfake content. To extract frame-level characteristics, we use Res-Next Convolutional Neural Networks. These attributes are then used to train a Recurrent Neural Network (RNN) based on Long Short-Term Memory (LSTM) to determine whether a video has been manipulated, i.e., whether it is a deepfake or a genuine video. We intend to train our deepfake detection model on a varied set of public datasets in order to improve its real-time performance. We improve the model's adaptability by learning features from different photos. Face-Forensic++, Deepfake Detection Challenge, and Celeb-DF datasets are used to extract videos. Furthermore, to assure competitive performance in real-world scenarios, our model will be assessed against a large amount of real-time data, including the YouTube dataset. Key Words: Temporal modelling, Deepfake technology , Long Short-Term Memory (LSTM) networks, Convolutional Neural Networks (CNNs)
32
Swetha, NG, Himasri Allu, Chandana KP Hari, Ujwal Kumar, Naga Sushwar, and BL Swathi. "Emotion Detection using Deep Learning CNN Model." International Journal of Microsystems and IoT 2, no. 9 (2024): 1187–96. https://doi.org/10.5281/zenodo.14099594.
Full textAbstract:
Facial Emotion Recognition (FER) is crucial in domains like human-computer interaction, mental health assessment, and marketing. This paper details the design and implementation of a FER model using Deep Convolutional Neural Networks (DCNNs) on the FER2013 dataset, which contains grayscale images labeled with seven emotions. Data augmentation and feature extraction are employed to enhance dataset diversity and reduce dimensionality. The DCNN architecture includes ReLU and Softmax activations for efficient non-linearity and multiclass classification, respectively, with Tanh and LeakyReLU showing promising results. The study explores the impact of pooling layers, identifying an optimal configuration of three layers. Hardware configurations significantly influence performance, with superior accuracy in System 2. Results highlight that balancing activation functions, pooling layers, and hardware specifications is key to optimizing CNN performance.
33
Ravindra, Vijay Eshi, Phade Gayatri, and Pawar Sushant. "Gesture Control Revolution: Enhancing Automotive Infotainment through Advanced Hand Gesture Recognition." Applied Science and Engineering Journal for Advanced Research 4, no. 2 (2025): 15–21. https://doi.org/10.5281/zenodo.15118126.
Full textAbstract:
In the ever-developing industry of automobiles, a focus should be made on the innovation of the car’s user experience while keeping the driver safe. The following paper therefore aims at proposing a new hand gesture recognition system to be implemented in car infotainment, which employs a modified CNN model enhanced with KNN for enhanced gesture mapping. The efficiency of the system was tested on a data of samples consisting of 10000 images of 10 different gestures performed by different users under different lighting conditions. The results obtained for the experimental evaluation proved that the used CNN reached the accuracy of 92,5% with the validation set and the further use of KNN for post-processing increased the classification accuracy up to 95,2%. Resource consumption was low, the CNN occupied roughly 50 MB of memory, that is why it is possible to use it for the in-vehicle system. A similar survey that targeted users showed that 85% of them were comfortable with the system as it was easy to learn and did not interfere with the control of infotainment functions. This research discusses the possibility of using gesture recognition technology to improve the user experience in vehicles making infotainment systems safer and more efficient.
34
Kothari, Sonali, Shwetambari Chiwhane, Shruti Jain, and Malti Baghel. "Cancerous brain tumor detection using hybrid deep learning framework." Indonesian Journal of Electrical Engineering and Computer Science 26, no. 3 (2022): 1651–61. https://doi.org/10.11591/ijeecs.v26.i3.pp1651-1661.
Full textAbstract:
Computational models based on deep learning (DL) algorithms have multiple processing layers representing data at multiple levels of abstraction. Deep learning has exploded in popularity in recent years, particularly in medical image processing, medical image analysis, and bioinformatics. As a result, deep learning has effectively modified and strengthened the means of identification, prediction, and diagnosis in several healthcare fields, including pathology, brain tumours, lung cancer, the abdomen, cardiac, and retina. In general, brain tumours are among the most common and aggressive malignant tumour diseases, with a limited life span if diagnosed at a higher grade. After identifying the tumour, brain tumour grading is a crucial step in evaluating a successful treatment strategy. This research aims to propose a cancerous brain tumor detection and classification using deep learning. In this paper, numerous soft computing techniques and a deep learning model to summarise the pathophysiology of brain cancer, imaging modalities for brain cancer, and automated computer-assisted methods for brain cancer characterization is used. In the sense of machine learning and the deep learning model, paper has highlighted the association between brain cancer and other brain disorders such as epilepsy, stroke, Alzheimer's, Parkinson's, and Wilson's disease, leukoaraiosis, and other neurological disorders.
35
M, Venkata Krishna Reddy, and Pradeep S. "Envision Foundational of Convolution Neural Network." International Journal of Innovative Technology and Exploring Engineering (IJITEE) 10, no. 6 (2021): 54–60. https://doi.org/10.35940/ijitee.F8804.0410621.
Full textAbstract:
Profound learning's goes to the achievement of spurs in a large number and understudies to find out about the energizing innovation. At this regular process of novices to venture the multifaceted nature of comprehension and applying profound learning. We present Convolution Neural Network (CNN) EXPLAINER, an intelligent representation instrument intended for non-specialists to learn and inspect (CNN)-Convolution Neural Network a fundamental profound learning model engineering. Our apparatus tends to key difficulties that fledglings face in finding out about Convolution Neural Network, it can be distinguish from pointing with educators and input with past understudies. Convolution Neural Network firmly incorporates representation outline that sums up the construction of CNN, and on-request, dynamic visual clarification sees that assist clients with understanding the hidden parts of CNNs. Constantly polished changes across levels of deliberation, our device empowers clients to examine the exchange between low-level numerical activities and undeniable level model designs. A subjective client study shows that Convolution Neural Network EXPLAINER helps clients all the more effectively comprehend the inward operations of CNNs, and is drawing in and agreeable to utilize. We additionally determine plan exercises from our examination. Created utilizing current web innovations, CNN EXPLAINER runs locally in clients' internet browsers without the requirement of establishment or particular equipment, widening the general preparation with current profound learning strategies.
36
Avinash, seekoli, Akkala Abhilasha, and Rathod Nagaraj. "A METHODOLOGY TO IDENTFY BRAIN TUMOR USING DEEP LEARNING TECHNIQUES." IJRSET FEBRUARY Volume 10 Issue 2 10, no. 2 (2023): 1–6. https://doi.org/10.5281/zenodo.8382717.
Full textAbstract:
Patients suffering from brain tumors are some of the most prevalent and aggressive, and in the latter stages of the disease, they have a very low life expectancy. The planning stage of surgical procedures is very important if the goal is to provide patients a higher quality of life throughout the course of their lives. Imaging techniques such as computed tomography (CT), magnetic resonance imaging (MRI), and ultrasound are often used in the process of locating malignancies in various parts of the body, including the brain, lungs, liver, breast, and prostate. In this particular instance, magnetic resonance imaging (MRI) scans are carried out in order to examine the patient's brain in search of signs of cancer. On the other hand, since an MRI gets so much information at once, it is difficult to differentiate between a tumor and something that isn't a tumor at the same time. This approach has a lot of drawbacks, the most notable one being that it can only produce accurate quantitative data for a constrained selection of photographs. There are also a great deal of additional restrictions associated with it. It is feasible that automated systems that can be relied on in a trustworthy manner might aid in the prevention of suicide. It is difficult to automatically classify brain tumors since the region and structure around a tumor may be somewhat variable. This is one reason why brain tumors can be so dangerous. In this article, fresh techniques to the early identification of malignant brain tumors are explained. CNNs are put to use in order to classify the data (Convolutional Neural Networks). According on the site of the tumor, this section classifies gliomas, meningiomas, pituitary tumors, and other types of tumors that are not malignant. The architectural design of the system's deeper levels is predicated on the use of tiny kernels as the building blocks. This is a reference to the very little amount of mass that the neuron has. The fact that CNN's accuracy in test results was 99.5% puts it in a class by itself above all other methods used by the present generation. In addition to this, it is simple to understand and much simpler to put into practice.
37
Nidhi, Gajimwar Ashmi Dahiwale Isha Walde Shyamal Dhabarde Prof. Monika Walde. "Automated Image Forgery Detection With Python." International Journal of Advanced Innovative Technology in Engineering 9, no. 3 (2024): 133–38. https://doi.org/10.5281/zenodo.12516039.
Full textAbstract:
Fake image detection has become increasingly important due to the widespread use of image editing software and the proliferation of fake images on social media and other online platforms. In this project, we propose a Python-based approach for detecting fake images using deep learning techniques. Our method involves preprocessing the images, extracting relevant features using convolutional neural networks (CNNs), and training a classifier to distinguish between real and fake images. We leverage state-of-the-art deep learning frameworks such as TensorFlow or PyTorch for model development and evaluation. Experimental results on benchmark datasets demonstrate the effectiveness of our approach in accurately identifying fake images. This project contributes to the ongoing efforts in combating misinformation and ensuring the authenticity of digital media content.
38
Solar, Mauricio, and Pablo Aguirre. "Deep learning techniques to process 3D chest CT." JUCS - Journal of Universal Computer Science 30, no. (6) (2024): 758–78. https://doi.org/10.3897/jucs.112977.
Full textAbstract:
The idea of using X–rays and Computed Tomography (CT) images as diagnostic method has been explored in several studies. Most of these studies work with slices of CT image in 2D, requiring less computational capacity and less time to process them than 3D. The processing of volumetric data (the complete CT images in 3D) adds an extra dimension of information. However, the magnitude of the data is considerably larger than working with slices in 2D, so extra computational processing is required. In this study a model capable of performing a classification of a 3D input that represents the volume of the CT scan is proposed. The model is able to classify the 3D input between COVID–19 and Non–COVID–19, but reducing the use of resources when performing the classification. The proposed model is the <em>ResNet–50</em> model with a new dimension of information added, which is a simple <em>autoencoder</em>. This <em>autoencoder </em>is trained on the same dataset, and a vector representation of each exam is generated and used together with the exams to feed the <em>ResNet–50</em>. To validate the proposal, the same proposed model is compared with and without the <em>autoencoder </em>module that provides more information to the proposed model. The proposed model obtains better metrics than the same model without the <em>autoencoder</em>, confirming that extracting relevant features from the dataset helps improve the performance of the model.
39
Varada, Venkata Sai Dileep, Rishitha Navuduru, Gummadi Rakesh, and Natarajan.P Prof. "DNA Sequencing using Machine Learning and Deep Learning Algorithms." International Journal of Innovative Technology and Exploring Engineering (IJITEE) 11, no. 10 (2022): 20–27. https://doi.org/10.35940/ijitee.J9273.09111022.
Full textAbstract:
<strong>Abstract:</strong> DNA Sequencing plays a vital role in the modern research. It allows a large number of multiple areas to progress, as well as genetics, meta-genetics, and phylogenetics. DNA Sequencing involves extracting and reading the strands of DNA. This research paper aims at comparing DNA Sequencing using “Machine Learning algorithms (Decision Trees, Random Forest, and Naive Bayes) and Deep Learning algorithms (Transform Learning and CNN)”. The aim of our proposed system is to implement a better prediction model for DNA research and get the most accurate results out of it. The “machine learning and deep learning models” which are being considered are the most used and reputed. A prediction accuracy of the higher range in deep learning is also being used which is also the better performer in different medical domains. The proposed models include “Decision Tree, Random Forest, Naive Bayes, CNN, and Transform Learning”. The Naive Bayes method gave greater accuracy of 98.00 percent in machine learning and the transform learning algorithm produced better accuracy of 94.57 percent in deep learning, respectively.
40
Thi, Ha Phan, Chung Tran Duc, and Fadzil Hassan Mohd. "Vietnamese character recognition based on CNN model with reduced character classes." Bulletin of Electrical Engineering and Informatics 10, no. 2 (2021): 962~969. https://doi.org/10.11591/eei.v10i2.2810.
Full textAbstract:
This article will detail the steps to build and train the convolutional neural network (CNN) model for Vietnamese character recognition in educational books. Based on this model, a mobile application for extracting text content from images in Vietnamese textbooks was built using OpenCV and Canny edge detection algorithm. There are 178 characters classes in Vietnamese with accents. However, within the scope of Vietnamese character recognition in textbooks, some classes of characters only differ in terms of actual sizes, such as “c” and “C”, “o” and “O”. Therefore, the authors built the classification model for 138 Vietnamese character classes after filtering out similar character classes to increase the model's effectiveness.
41
Arsha, Anish, Shouckath Kolliyath Shebin, Joshy Sneha, K. Athira, and A. S. Revathy. "DEEP FAKE DETECTION USING MACHINE LEARNING." Research and Reviews: Advancement in Cyber Security 1, no. 3 (2024): 23–29. https://doi.org/10.5281/zenodo.13382382.
Full textAbstract:
<em>The recent proliferation of free, deep learning-based tools has democratized the creation of high-fidelity “deepfake” videos, where facial exchanges exhibit minimal manipulation artifacts. While advancements in visual effects have facilitated video manipulation for decades, deep learning has ushered in an era of unprecedented realism and accessibility for generating such “AI-synthesized media.” While crafting deepfakes is now relatively straightforward, their detection remains a significant challenge due to the complexities involved in training algorithms to recognize these subtle manipulations. This paper presents a novel deepfake detection method employing a combined convolutional neural network (CNN) and recurrent neural network (RNN) architecture. The CNN extracts frame-level features, and these features are subsequently fed into an RNN trained to classify videos as manipulated or authentic. The RNN’s ability to learn temporal dependencies allows it to effectively detect inconsistencies between frames, often introduced by deepfake creation tools. We evaluate our system on a comprehensive dataset of synthetic videos and demonstrate competitive performance, highlighting the potential of our relatively simple architecture in addressing this critical challenge.</em>
42
Satyanegara, Hartina Hiromi, and Kalamullah Ramli. "Implementation of CNN-MLP and CNN-LSTM for MitM Attack Detection System." Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi) 6, no. 3 (2022): 387–96. http://dx.doi.org/10.29207/resti.v6i3.4035.
Full textAbstract:
Man in the Middle (MitM) is one of the attack techniques conducted for eavesdropping on data transitions or conversations between users in some systems secretly. It has a sizeable impact because it could make the attackers will do another attack, such as website or system deface or phishing. Deep Learning could be able to predict various data well. Hence, in this study, we would like to present the approach to detect MitM attacks and process its data, by implementing hybrid deep learning methods. We used 2 (two) combinations of the Deep Learning methods, which are CNN-MLP and CNN-LSTM. We also used various Feature Scaling methods before building the model and will determine the better hybrid deep learning methods for detecting MitM attack, as well as the feature selection methods that could generate the highest accuracy. Kitsune Network Attack Dataset (ARP MitM Ettercap) is the dataset used in this study. The results prove that CNN-MLP has better results than CNN-LSTM on average, which has the accuracy rate respectively at 99.74%, 99.67%, and 99.57%, and using Standard Scaler has the highest accuracy (99.74%) among other scenarios.
43
Ms. Jyoti Pandurang Kshirsagar. "Leveraging Faster CNN (F-CNN) for Effective Breast Cancer Classification." Advances in Nonlinear Variational Inequalities 28, no. 2 (2024): 117–35. http://dx.doi.org/10.52783/anvi.v28.1855.
Full textAbstract:
Within the scope of this work, a novel classification method for the diagnosis of breast cancer that is based on deep learning is also described. In this particular instance of breast cancer, which is the most common form of cancer in females, early detection is absolutely necessary in order to get better treatment outcomes. Notwithstanding their effectiveness, traditional diagnostic techniques have drawbacks such high expenses and possible errors. The high dimensionality and instability in tumor morphology that are particular problems with breast cancer imaging are intended to be addressed by the suggested techniques. Using publically available datasets for rigorous training and validation, a bespoke deep learning model is designed and implemented, and an extensive evaluation of current deep learning methodologies is conducted as part of the research. The model's accuracy and resilience are significantly improved when compared to the performance of existing classification algorithms. To enhance diagnosis accuracy in the characterization of breast cancer, this study makes utilizes of deep learning, more especially faster convolutional neural networks. The investigation also looks at the model's clinical usefulness, providing information about how it might be incorporated into diagnostic procedures. According to the findings, it appears that the highlighted methodology has the potential to transform the diagnosis of breast cancer by offering a dependable and automated solution that can improve early detection and patient outcomes. In just 3 epochs, the model obtained a remarkable accuracy of 92% on DDSM dataset.
44
Smithu, B. S., D. R. Janardhana, C. P. Leela, and G. Pushpa. "Forest Fire Risk Assessment and Detection using Deep Learning Models." Indian Journal Of Science And Technology 17, no. 46 (2024): 4921–28. https://doi.org/10.17485/ijst/v17i46.2138.
Full textAbstract:
Background: There is a severe need to detect any kind of fire in a faster and accurate method, especially forest fires to stop huge losses to the human community and the environment losses. The main purpose of the proposal is to identify and evaluate the accuracy of the existing Artificial Intelligence (AI) methods for detecting fire and improve the methods to detect fire in real-world scenarios in faster and accurate methods. Methods: The proposal uses a dataset to train a model, and in addition uses a few test images from an existing database to test the models developed. We develop and test the following neural network models namely, Deep Neural Network (DNN), Recurrent Neural Network (RNN), Convolution Neural Network (CNN), and a combination of RNN and CNN. Findings: A conservative estimate of the yearly losses caused by forest fires in India for the entire nation is 440 crores INR. The loss of biodiversity, soil moisture, nutrients, and other intangible advantages is not factored in this assessment. The proposed models namely DNN, RNN, CNN, and RNN+CNN give an accuracy of 55%, 61%, 55% and 98% respectively. Novelty and applications: The RNN+CNN model proposed to have good accuracy which is much better compared to existing models. The model in addition can be used in real-time CCTV surveillance which can predict the fire in real-time with a faster alert duration of less than 30 sec. Keywords: Fire detection, CNN, RNN, DNN, Artificial Intelligence (AI), Environment loss, Fire Losses
45
Ahmad, AL Smadi, Mehmood Atif, Abugabah Ahed, Almekhlafi Eiad, and Mohammad Al-smadi Ahmad. "Deep convolutional neural network-based system for fish classification." International Journal of Electrical and Computer Engineering (IJECE) 12, no. 2 (2022): 2026–39. https://doi.org/10.11591/ijece.v12i2.pp2026-2039.
Full textAbstract:
In computer vision, image classification is one of the potential image processing tasks. Nowadays, fish classification is a wide considered issue within the areas of machine learning and image segmentation. Moreover, it has been extended to a variety of domains, such as marketing strategies. This paper presents an effective fish classification method based on convolutional neural networks (CNNs). The experiments were conducted on the new dataset of Bangladesh’s indigenous fish species with three kinds of splitting: 80-20%, 75-25%, and 70-30%. We provide a comprehensive comparison of several popular optimizers of CNN. In total, we perform a comparative analysis of 5 different state-of-the-art gradient descent-based optimizers, namely adaptive delta (AdaDelta), stochastic gradient descent (SGD), adaptive momentum (Adam), adaptive max pooling (Adamax), Root mean square propagation (Rmsprop), for CNN. Overall, the obtained experimental results show that Rmsprop, Adam, Adamax performed well compared to the other optimization techniques used, while AdaDelta and SGD performed the worst. Furthermore, the experimental results demonstrated that Adam optimizer attained the best results in performance measures for 70-30% and 80-20% splitting experiments, while the Rmsprop optimizer attained the best results in terms of performance measures of 70-25% splitting experiments. Finally, the proposed model is then compared with state-of-the-art deep CNNs models. Therefore, the proposed model attained the best accuracy of 98.46% in enhancing the CNN ability in classification, among others.
46
K, Gayathri, and Thangavelu S. "Novel deep learning model for vehicle and pothole detection." Indonesian Journal of Electrical Engineering and Computer Science 23, no. 3 (2021): 1576–82. https://doi.org/10.11591/ijeecs.v23.i3.pp1576-1582.
Full textAbstract:
The most important aspect of automatic driving and traffic surveillance is vehicle detection. In addition, poor road conditions caused by potholes are the cause of traffic accidents and vehicle damage. The proposed work uses deep learning models. The proposed method can detect vehicles and potholes using images. The faster region-based convolutional neural network (CNN) and the inception network V2 model are used to implement the model. The proposed work compares the performance, accuracy numbers, detection time, and advantages and disadvantages of the faster region-based convolution neural network (Faster R-CNN) with single shot detector (SSD) and you only look once (YOLO) algorithms. The proposed method shows good progress than the existing methods such as SSD and YOLO. The measure of performance evaluation is Accuracy. The proposed method shows an improvement of 5% once compared with the previous methods such as SSD and YOLO.
47
Kanagala, Hari Krishna, and Dr V. V. Jayarama Krishnaiah. "Enhanced Mechanism for Classification of Glaucoma Images Using Deep Learning based CNN." Journal of Advanced Research in Dynamical and Control Systems 11, no. 10 (2019): 111–18. http://dx.doi.org/10.5373/jardcs/v11i10/20193013.
Full text
48
Lin, Tianying, Ang Liu, Xiaopei Zhang, et al. "Analyzing OAM mode purity in optical fibers with CNN-based deep learning." Chinese Optics Letters 17, no. 10 (2019): 100603. http://dx.doi.org/10.3788/col201917.100603.
Full text
49
Dhrishya, Suresh, Libi Kajal, S. Pradeep Neeraja, B. Rajalakshmy, and A. Sidhik. "Unsound: Software to Recognize Sign Language Using Deep Learning." Journal of Advancement in Software Engineering and Testing 7, no. 1 (2023): 8–16. https://doi.org/10.5281/zenodo.10184198.
Full textAbstract:
<i>The extraction of complicated hand gestures with continually changing shapes for sign language identification is regarded as a difficult task in computer vision. This research proposes the recognition of American sign language movements using convolutional neural networks (CNN), a strong artificial intelligence technology. The dataset contains 87,000 images, each sized 200x200 pixels. It's designed for a classification task, with 29 different categories. These categories include the letters A-Z (26 classes), as well as three additional classes for SPACE, DELETE, and NOTHING. The goal is to train a machine learning model to correctly categorize these images into their respective classes. CNN training is carried out with a variety of sample sizes, each of which includes many sets of individuals and viewing angles. To improve recognition accuracy, various CNN architectures were created and tested using our sign language data. Alongside our sign language recognizing software (UNSOUND) we have also incorporated the conversion of text to sign language for the convenience of those who are not familiar with sign language.</i>
50
Nehal, Mohamed Ali, Mostafa Abd El Hamid Marwa, and Youssif Aliaa. "Sentiment Analysis for Movies Reviews Dataset Using Deep Learning Models." International Journal of Data Mining & Knowledge Management Process (IJDKP) 9, no. 2/3 (2019): 19–27. https://doi.org/10.5281/zenodo.3340668.
Full textAbstract:
Due to the enormous amount of data and opinions being produced, shared and transferred everyday across the internet and other media, Sentiment analysis has become vital for developing opinion mining systems. This paper introduces a developed classification sentiment analysis using deep learning networks and introduces comparative results of different deep learning networks. Multilayer Perceptron (MLP) was developed as a baseline for other networks results. Long short-term memory (LSTM) recurrent neural network, Convolutional Neural Network (CNN) in addition to a hybrid model of LSTM and CNN were developed and applied on IMDB dataset consists of 50K movies reviews files. Dataset was divided to 50% positive reviews and 50% negative reviews. The data was initially pre-processed using Word2Vec and word embedding was applied accordingly. The results have shown that, the hybrid CNN_LSTM model have outperformed the MLP and singular CNN and LSTM networks.