Search

Sign language is a crucial mode of communication for individuals who are deaf or hard-of-hearing, and a reliable sign language recognition system can greatly aid in bridging the communication gap. According to statistics provided by the Government of India, there are only around 250 accredited sign language interpreters in India, servicing a deaf population of 1.8 million to 7 million individuals. This paper presents a real-time sign language recognition system for Indian Sign Language (ISL) using YOLOv8 for object detection. The system uses a single convolutional neural network (CNN) architecture to process the entire image in one forward pass and predict the bounding boxes and class probabilities for each sign in the image. The results of the study demonstrate the feasibility and potential of the system in revolutionizing communication for the deaf and hard of hearing community and have applications in gesture-controlled robotics, home automation, game control, human-computer interaction, and sign language interpretation. [ABSTRACT FROM AUTHOR]

After a deep questioner among students, faculty members and higher education experts. The demand for qualified engineers in the specific field as understandably gone down. The situation is grimmer for tier-2 and tier-3 colleges in India. The purpose of opting engineering course is not to get expertise in the particular course, instead to get a job in government sector with valid degree. The problem identified in the field of engineering education towards improving the quality in education is addressed with the help of exploratory data analysis. The dataset used in our experiment is collect from UCI machine Learning Repository having 33 variables and 1044 observations. The contribution made in the paper is to identify the vital attributes using single and multi-variant regression techniques. To perform prediction using Decision Tree, Random Forest, Support vector Machine and compare their performance in terms of classification accuracy and F-Score. In addition to that a convolution Neural Network (CNN) model is established in which, the vital attributes identified using regression techniques are provided as inputs and weights at each stage is estimated using Gradient Decent algorithm with step size 0.5. The classification accuracy from 63% is improved to 97% with the help of CNN model in 26664 iterations. The finding of the present research states that student having backlogs are less frequently opting for higher studies. [ABSTRACT FROM AUTHOR]

The main factor causing blindness in countries like India and others is Diabetic Retinopathy. Millions of people around the world are affected by DR, which causes vision loss and blindness. Detecting diabetes mellitus early plays an essential role in preventing the loss of vision that can result from diabetes mellitus being left untreated for a considerable amount of time. It is becoming increasingly more important to detect diabetic retinopathy early through automated systems instead of manual screening procedures like fluorescein angiography, optical coherence tomography etc., as it can lead to blindness among patients with uncontrolled diabetes. There are several published studies on machine learning and deep learning-based DR detection systems. In this paper, we explore the basics of advanced AI technologies used in DR analysis and early detection. The purpose of this study is to review DR detection methods are examined from a variety of perspectives. It includes fundus datasets, pre-processing of images, methodologies, machine learning and deep learning-based methodologies, and performance metrics. As well as presenting the review findings, it also presents the authors' observations. There were numerous public datasets for DR detection that are readily available. The Artificial Neural Network proved to be the better classifier over other machine learning methods based on shape, texture, and statistical features for DR detection. The purpose of this study is to propose the Convolutional Neural Network (CNN) through GoogLeNet for betterment based on the previous CNN model which was implemented using deep learning. Furthermore, the scientific community dedicated to developing automated DR investigative techniques. This review offers a thorough overview of DR detection methods. [ABSTRACT FROM AUTHOR]

In recent years, the human population has reached its peak in both the developing countries and developed countries like India and China. An increase in personnel density ended up being a hard task to maintain or manage people flow in socio-public places. However, people flow management is out of control for humans, which may also cause some safety damages. This paper aims to give an optimal solution for people flow with the help of machine learning, deep learning, and also AI-cameras are used for counting the number of people entering and leaving a user-defined place. Here, also the centroid tracking and centroid detecting algorithms are used to count the number of people together with their direction. This technique helps to replace manual security,in-store traffic management, and much more. [ABSTRACT FROM AUTHOR]

India is an Agriculture based country and about 60% of the population depends on agriculture for their living. But there is a substantiable reduction in agricultural production has happened due to unstable climate conditions and global warming. Indian farmers are always lacking the support of advanced technologies for improving and maximizing their yield. If they get explicit information about soil type, nutrients, pH value, changes in climatic factors, previous year yield etc., they can optimize their work and yield. With the advancement of machine learning, big data analytics and cloud computing technologies, the climate and crop yield can be got predicted to the farmers. Prediction of yield in advance can help the farmers to take corrective decisions about fertilization, storage and marketing to increase their production and revenue. A huge number of studies have been conducted worldwide, on agriculture sector for the prediction of crop yield, plant diseases etc. This paper presents a brief study of different researches in India on crop yield prediction based on Machine Learning and deep learning. Most of the researchers used environmental parameters like temperature, rainfall and soil type as the main features for prediction. A number of machine learning algorithms are employed in this field and it is really a complex task to identify the best algorithm. The researchers tried to accurately predict the crop yield and thereby suggesting ways to improve production and efficient use of fertilizers. [ABSTRACT FROM AUTHOR]

Time series forecasting is a decisive step in data modeling and a significant area in machine learning. This paper presents Long short-term memory (LSTM) network, a deep learning neural network for predicting Covid-19 cases in India. The neural network models are trained and tested with Covid-19 case data sets obtained from PRS Legislative Research database. Further, the parameter optimization is carried out for choosing the optimal network. The parameters considered for evaluating the performance of LSTM network are RMSE, number of epochs, accuracy and loss. The results are compared with various recurrent neural network models and autoregressive model. The results revealed an improved accuracy of 92.8% for LSTM network in predicting the transmission of Covid-19 in India. [ABSTRACT FROM AUTHOR]

Every year, tens of thousands of people are declared missing in India. A significant number of people who have gone missing have yet to be found. With the aid of facial recognition, this paper suggests a unique deep learning technique application for recognizing the identified missing person from the images of a large number of people available. Photographs of suspicious people may be uploaded to a shared portal, including location and timings. The snapshot will be linked to the missing person's images in the repository. The input individual picture is classified, and the picture with the good fit is chosen from a missing person database. Using the image reported by the public, an algorithm is learned such that we can accurately find missing person in archive with missing people's images generated. Face recognition is done using Convolution Neural Network (CNN) which is a common deep learning algorithm to process image-based applications. Face features are derived from pictures using a VGG(Visual Geometry Group)-Face deep architecture CNN model that has been pre-trained. Unlike other deep learning algorithms, ours only uses a convolutional network as a high feature attribute, with the qualified KNN classifier doing the individual recognition. When properly prepared, VGG-Face, the highest performing CNN algorithm for face detection, outputs a deep learning algorithm that is distortion, lighting, and person age and exceeds previous approaches in facial recognition-based missing person identification. This recognition device classification accuracy is 90%. [ABSTRACT FROM AUTHOR]