Recognizing human activity is a challenging task in many applications like visual surveillance, Human–computer interaction, etc. Numerous machine learning (ML) and deep learning (DL) approaches are used to emulate the human behavior of everyday life, as they provide better recognition by learning complex features. This paper proposes an innovative human activity recognition framework with the following steps. Initially, the given video is converted into a set of frames and then applies median filtering to remove the noise for efficient recognition. Subsequently, the pixel-relatedness retrieval-fuzzy clustering means algorithm (PRR-FCM) approach is proposed for image segmentation, in which a pixel-relatedness retrieval strategy is used that considers only the proximity-related pixel for efficient segmentation. The retrieval of features is an important aspect, thereby this paper goes with the extraction of texture features, relative standard deviation induced in multi-texon (RSIMT), in which relative standard deviation is induced for interchanging the center pixel value. Also, the other features like the hierarchy of the skeleton and local binary pattern (LBP) are extracted from the segmented image. Further, the retrieved features are fed to the hybrid model that combines models like modified deep Maxout (MDM) and convolution neural network (CNN) models to recognize human activity efficiently. The modified exponential linear unit (ELU) activation function and Log loss function are the enhancing formulation of the MDM model to determine accurate recognition results.

Glaucoma is currently one of the most significant causes of permanent blindness. Fundus imaging is the most popular glaucoma screening method because of the compromises it has to make in terms of portability, size, and cost. In recent years, convolution neural networks (CNNs) have revolutionized computer vision. Convolution is a “local” CNN technique that is only applicable to a small region surrounding an image. Vision Transformers (ViT) use self-attention, which is a “global” activity since it collects information from the entire image. As a result, the ViT can successfully gather distant semantic relevance from an image. This study examined several optimizers, including Adamax, SGD, RMSprop, Adadelta, Adafactor, Nadam, and Adagrad. With 1750 Healthy and Glaucoma images in the IEEE fundus image dataset and 4800 healthy and glaucoma images in the LAG fundus image dataset, we trained and tested the ViT model on these datasets. Additionally, the datasets underwent image scaling, auto-rotation, and auto-contrast adjustment via adaptive equalization during preprocessing. The results demonstrated that preparing the provided dataset with various optimizers improved accuracy and other performance metrics. Additionally, according to the results, the Nadam Optimizer improved accuracy in the adaptive equalized preprocessing of the IEEE dataset by up to 97.8

The most common cause of blindness in the world is glaucoma. It can lead to a decline in eyesight and quality of life if not addressed within the stipulated time. In computer vision, convolutional neural networks (CNNs) have achieved a viable solution. The most standard glaucoma screening method is fundus imaging due to the compromises it makes between mobility, size, and price. In this paper, we utilized the RIM-1 DL dataset in two different versions: RIM-1 Hospital and RIM-1 Randomly. To enhance the image quality and increase accuracy, we used median filtering pre-processing techniques in this work. We also used fundus images to automatically diagnose glaucoma using six different Image-net-trained models (DenseNet121, DenseNet169, VGG19, InceptionV3, ResNet50, and MobileNetV2). The six CNN architectures mentioned above are modified using normalization, inception, and global pooling layers in the model with the pre-trained classification models. Results indicate that the DenseNet-121 is the most accurate model in this work, improving accuracy in the RIM-1 Hospital dataset from 70.68 to 80.13 and in the RIM-1 randomly generated dataset from 79.31 to 87.67 after the models were fine-tuned.

Human Activity Recognition (HAR) is an AI-based technique aimed at identifying and labeling human activities. It gathers activity data from sources like wearable sensors and smart devices. Essentially, it categorizes human activities based on interactions, movements, and actions. Currently, it finds widespread application across various domains, leading to continuous research efforts for practical improvements. The present study revolves around HAR, a hybrid DM-CNN model, where the activities of humans are recognized by four phases including preprocessing, segmentation, feature extraction, and recognition. Here, the hybrid DM-CNN model uses a specific set of extracted features for enhanced training and they are MTH, IGBP and hierarchy of skeleton. Most importantly, the hybrid DM-CNN model provides excellent results on human activity recognition through the development of the Modified Osprey Optimization (MOO) algorithm for fine-tuning the DM and CNN models. The specificity of the proposed model is 95

Summary Classifying human actions from still images or video sequences is a demanding task owing to issues, like lighting, backdrop clutter, variations in scale, partial occlusion, viewpoint, and appearance. A lot of appliances, together with video systems, human–computer interfaces, and surveillance necessitate a compound action recognition system. Here, the proposed system develops a novel scheme for HAR. Initially, filtering as well as background subtraction is done during preprocessing. Then, the features including local binary pattern (LBP), bag of the virtual word (BOW), and the proposed local spatio‐temporal features are extracted. Then, in the recognition phase, an ensemble classification model is introduced that includes Recurrent Neural networks (RNN 1 and RNN 2) and Multi‐Layer Perceptron (MLP 1 and MLP 2). The features are classified using RNN 1 and RNN 2, and the outputs from RNN 1 and RNN 2 are further classified using MLP 1 and MLP 2, respectively. Finally, the outputs attained from MLP 1 and MLP 2 are averaged and the final classified output is obtained. At last, the superiority of the developed approach is proved on varied measures.

Glaucoma is a chronic, irrevocable eye ailment in which the optic nerve slowly deteriorates, impairing vision and quality of life. Specialists can diagnose clinical glaucoma; however, the procedures are either expensive or time-consuming. In this paper, we have proposed an enhanced version of various pre-trained CNN models (DenseNet, ResNet, VGG, Inception, and MobileNet) using batch normalization, inception module, and global pooling with the Swish activation function. We have used three datasets (ACREMA, LAG, and IEEE Fundus images) and pre-processed versions of these datasets in this work. Moreover, we have also proposed a hybrid region of interest (ROI) extraction approach in which we have extracted the region of interest using a combination of the Brightest Spot algorithm and Hough Circles implementation for pre-processing of the fundus images to improve the overall performance. We have also applied manual hyperparameter optimization for learning rate, batch size, SGD optimizer, and fine-tuning step. The proposed framework acquired an improved accuracy of 98–99

Glaucoma is the foremost cause of permanent loss of visioahn; it develops slowly with no discernible sign. Early glaucoma recognition is critical because it might aid to slow the progression of the disease. Customary schemes are less precise and manual. As a result, automatic glaucoma analysis is required to identify glaucoma at the earliest with greater precision. The purpose here is to bring in a new scheme in which pre-processing is done using Gaussian filtering, which aids in the removal of unwanted noise in images. Then Optic Cup segmentation is performed using the Modified Level Set Algorithm. Followed by segmentation, the morphological features (disc area, cup area, and blood vessel), as well as non-morphological features (Color, Shape, and Modified LBP), are derived. The Blood vessel thickness is 5 to 100 micrometers. These features are then classified using the Opti-mized CNN framework, where the weights get optimized via the Self Adaptive Butterfly Optimization Algorithm (SA-BOA). The precision of the developed approach was 21.16%, 7.35%, 6.62%, 2.98%, 4.29%, 3.89%, 5.67%, 6.23%, 6.79%, and 1.63% better than the values obtained for conservative techniques Similarly, the adopted model's negative metrics show negligible values when compared to other models. Thus, the proposed method supremacy was validated successfully.

Nowadays Electronic Health Care System is growing to a large extent. As data retrieve from IoT devices, that data is needed to store in a database as the different patients have different data. Due to a lack of intrinsic security safeguards, the Internet of Things is prone to privacy and security breaches. The information gathered from IoT devices is mostly saved in a database This data is very important for a particular patient. If this data is altered by any intruder, then the doctor cannot find the actual problem of the patient and also cannot give the patient proper treatment, as a result, it causes great harm to the patient. As a result, a security mechanism for the data contained in the database is required. Blockchain technology will help a lot in this situation. We make an IoT-based prototype that uses Blockchain technology to get rid of this anonymous data access, in the common word patient's data are private through this system.

Glaucoma is a fast-growing retinal disease, which is a chronic and neurodegenerative disease that affects patients progressively. The neuro-retinal nerve called the optic nerve, which connects the eye to the brain damages the eye due to lack of proper diagnosis. To diagnose this disease, ophthalmologists need to perform various tests. However, taking more tests requires a lot of time and that is expensive. Also, the early stage of glaucoma may not show any vision loss to the affected people. For detecting glaucoma, this paper intends to propose the glaucoma detection model that follows four phases including preprocessing, segmentation, Feature extraction, and glaucoma detection. In the Preprocessing phase, the retinal images are processed by median filtering. Subsequently, subjecting the preprocessed image to the segmentation, where the optic cup will be segmented by the improved U-net model. Subsequently, the features such as cup-to-disc ratio, inferior superior nasal temporal features, fractal features, and proposed Local Gabor features are derived from the segmented image. These extracted features are trained to the optimized DeepMaxout classifier for detection purposes. Finally, a novel self-improved Beluga whale optimization algorithm is proposed for the optimal training of DMN by tuning its optimal weights.

INTRODUCTION: Nowadays one of the primary causes of permanent blindness is glaucoma. Due to the trade-offs, it makes in terms of portability, size, and cost, fundus imaging is the most widely used glaucoma screening technique. OBJECTIVES:To boost accuracy,focusing on less execution time, and less resources consumption, we have proposed a vision transformer-based model with data pre-processing techniques which fix classification problems. METHODS: Convolution is a “local” technique used by CNNs that is restricted to a limited area around an image. Self-attention, used by Vision Transformers, is a “global” action since it gathers data from the whole image. This makes it possible for the ViT to successfully collect far-off semantic relevance in an image. Several optimizers, including Adamax, SGD, RMSprop, Adadelta, Adafactor, Nadam, and Adagrad, were studied in this paper. We have trained and tested the Vision Transformer model on the IEEE Fundus image dataset having 1750 Healthy and Glaucoma images. Additionally, the dataset was preprocessed using image resizing, auto-rotation, and auto-adjust contrast by adaptive equalization. RESULTS: Results also show that the Nadam Optimizer increased accuracy up to 97% in adaptive equalized preprocessing dataset followed by auto rotate and image resizing operations. CONCLUSION: The experimental findings shows that transformer based classification spurred a revolution in computer vision with reduced time in training and classification.

Human Activity Recognition is a recognition technique which identify and name the human activities by AI. It collects activity data from devices such as wearable sensors and smart devices etc. Literally, it classifies human activities on the basis of interaction, motion and action. Nowadays, its usability is spread in various applications so, the research on Human Activity Recognition is still ongoing with viable enhancements. Though, this technique needs to solve limitations such as collection and processing of data, stability of hardware and approaches, complexity in activity detection and messed up activities. This work is based on Human Activity Recognition which is trained by the hybrid model of classifiers such as LSTM and Bi-GRU via proposed Self-Improved Namib Beetle Optimization model (SINBO). The activity recognition is detected by two stages: (1) Preprocessing, (2) Feature Extraction and (3) Classification. The extracted features areImproved bag of visual words, Local texton XOR pattern and SLIF are given as an input in classification stage. The hybrid model of LSTM and Bi-GRU is enhanced by optimal tuning of weights via SINBO algorithm.

A wide variety of uses, such as video interpretation and surveillance, human-robot interaction, healthcare, and sport analysis, among others, make this technology extremely useful, human activity recognition has received a lot of attention in recent decades. human activity recognition from video frames or still images is a challenging procedure because of factors including viewpoint, partial occlusion, lighting, background clutter, scale differences, and look. Numerous applications, including human-computer interfaces, robotics for the analysis of human behavior, and video surveillance systems all require the activity recognition system. This work introduces the human activity recognition system, which includes 3 stages: preprocessing, feature extraction, and classification. The input video (image frames) are subjected for preprocessing stage which is processed with median filtering and background subtraction. Several features, including the Improved Bag of Visual Words, the local texton XOR pattern, and the Spider Local Picture Feature (SLIF) based features, are extracted from the pre-processed image. The next step involves classifying data using a hybrid classifier that blends Bidirectional Gated Recurrent (Bi-GRU) and Long Short Term Memory (LSTM). To boost the effectiveness of the suggested system, the weights of the Long Short Term Memory (LSTM) and Bidirectional Gated Recurrent (Bi-GRU) are both ideally determined using the Improved Aquila Optimization with City Block Distance Evaluation (IACBD) method. Finally, the effectiveness of the suggested approach is evaluated in comparison to other traditional models using various performance metrics.

This research work provides a systematic review on blockchain-based applications present across multiple domains. Also, a comprehensive classification of technology applications across diverse sectors such as healthcare, supply chain, education, automobile industry, etc. has been presented. The twenty-first century has introduced significant innovations and discoveries that have completely changed our lives. In this way, Computers were first introduced followed by the internet, internet of things, and this has recently resulted in the introduction of "Blockchain" technology. This technology is considered as the most important innovation since the internet, and it is also known as the future internet since it creates a decentralized network chain. It has huge potential to impact, disrupt, and change human activities. Also, blockchain becomes an ideal tool for creating a trust-based solution. Nowadays several private, public sector firms, and government organizations are initiating new research work in the blockchain domain. This paper discusses innovative ideas, technologies, and its real-time applications by considering different industries, where blockchain can be applied to create new business opportunities.

With the recent advancement of technology in this era, software industries has grown outrageously. Software industry has shown such great hike in technology which is non-comparable to any other industries. Various methods have been established which improves the software quality one such method is Agile. Agile software development has gained a lot of attention because of its simplicity and ease of use. Agile software development is an approach which produces quality software with remarkable team interaction and more of customer involvement. Agile method is basically ideally suited for a scenario where requirements are changing in continuous manner. One of the most important advantage of using Agile is, it takes less time for software release, easy to understand and require less documentation. This paper deals with various agile methods, their comparison, advantages, shortcomings and XCRUMBAN, a new proposed framework to overcome those mentioned shortcomings.

The delivery of Cloud computing is a method for delivering information /services in which resources are retrieved from the Internet through web-based tools and applications, as opposed to a direct connection to a server rather than keeping files on a proprietary hard drive or local storage device.We have studied a lot of algorithm for reducing the response time of load balancing algorithm in cloud computing environment.We also measured the execution time of different algorithms.In this paper we will compare execution time, processing time of data center, response time of various algorithms.

The audio and video synchronization plays an important role in speech recognition and multimedia communication.The audio-video sync is a quite significant problem in live video conferencing.It is due to use of various hardware components which introduces variable delay and software environments.The objective of the synchronization is used to preserve the temporal alignment between the audio and video signals.This paper proposes the audio-video synchronization using spreading codes delay measurement technique.The performance of the proposed method made on home database and achieves 99% synchronization efficiency.The audio-visual signature technique provides a significant reduction in audio-video sync problems and the performance analysis of audio and video synchronization in an effective way.This paper also implements an audio-video synchronizer and analyses its performance in an efficient manner by synchronization efficiency, audio-video time drift and audio-video delay parameters.The simulation result is carried out using mat lab simulation tools and simulink.It is automatically estimating and correcting the timing relationship between the audio and video signals and maintaining the Quality of Service.

At present, internet has become part of life in human beings life. Internet is now a gigantic library that is composed of documents, files, images, videos, content and websites. Unlimited amount of data is added regularly to the library from different mediums and in different formats. Cloud computing is associated with internet computing. Cloud computing has undoubtedly benefited both service provider and clients in great extent. There is rapid increase in cloud‘s customers constantly. Although, the cloud data centers comprised of tremendous power but due to expeditious requests of users there is sudden need of balancing load. However, load balancing emerged as the conspicuous issue in the cloud heterogeneous environment. A comprehensive literature survey on various load balancing algorithms is presented here which addresses that there can be reduction in response time and data center request processing time by using efficient load balancing policies.

There are numerous Model Driver Engineering (MDE) methodologies but Object Management Group (OMG) approved of Model Driver Architecture (MDA). MDA methodology has a target to systemize the software progressing procedure with the use of models rather than the old-fashioned coding based on isolation of the related theory. During the month of June in the year 2014, OMG brought second edition of MDA guide into the market in attempt to understand about essential values and to back first edition of MDA guide which came out in 2003 and had thorough provisions included within. An interval of 11 years allows the investigators to come out of behind and put forward their viewpoint with the various clarifications of MDA provisions. People often gets mistaken and consumed about what is outside of MDA scope and what is inside it. Severely mentioning to MDA standard (not MDE in general), a review of present MDA Literature is given by us here. A bit of a spotlight is also cast upon the MDA research directions, more particularly upon mechanizations of MDA progress procedure and the raised areas which it aims.

Blogs, comments and reviews have now become an integral part of people who want to read them in order to be informed regarding other people opinion.This helps them to gain an overview of what other people say so that they might take a decision based on other people recommendation.Most of the time the user may not been in a position to read all the opinions and then take an informed decision about the product or services which he/she wants to take.Also it has been seen that most of the websites use different approach like star rating, numerical rating, to depict the information to the people who want to read the reviews.In this paper our aim is to develop a system for providing a method to help and explore good restaurants and specific dishes which a user wants to know based on past experiences of the people.The basic approach is to extract opinions from the websites and to extract the meaning of those sentences by applying Natural Language Processing techniques and then give the rating on a 5-point scale.