Volume 12 | Issue 5 |

 Abstract

The goal of this project is to promote inclusivity in social interactions and employment possibilities by addressing the communication issues that people who are blind, deaf, or mute confront. The suggested remedy consists of a two-way smart communication system made to make it easier for people with and without sensory impairments to communicate with each other. The initial component of the system provides an easy-to-use interface that helps people who are blind, deaf, or mute effectively communicate. To empower users with a range of sensory needs, the system makes use of cutting-edge technologies including gesture control, speech recognition, and haptic feedback. Voice instructions, tactile gestures, or both can be used by users to input messages. The system then converts these inputs into a textual and auditory format that is simple to comprehend, making the messages transmitted understandable to people with different levels of communication proficiency. With the use of machine learning algorithms, this interface enables accurate and speedy transcription of spoken words into text through speech-to-text conversion.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Inclusivity, Social Interactions, Employment opportunities, sensory impairments, gesture control, speech recognition, haptic feedback, Voice instructions, tactile gestures, user-friendly interface, machine learning algorithms, speech-to-text conversion

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Textualizing Sign Language, the challenge purpose is to have a look at numerous techniques for effective inter-communication between Sign Language and its impact on communication. English Language. The version can identify almost every alphabet. Various gestures which include each palm have been added to our dataset. This model has big potential as it could interpret any gesture of diverse Sign Languages if provided in the dataset. The consumer can also upload extra gestures within the dataset, making it rather custom-designed. Further, the data exists to an application to convert the obtained textual content to speech.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Early Detection, Deep learning, CNN Algorithm, Image Processing.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The utilization of deep learning techniques in "Deep Learning-driven Food Recognition and Calorie Estimation for Intelligent Diet Monitoring" offers a unique method to improve diet monitoring and encourage healthy eating habits. The main goal of this initiative is to accurately identify various food items and estimate their calorie content in real-time, granting users access to intelligent and tailored diet monitoring features. By utilizing the Python programming language and the MobileNet architecture model for food recognition and calorie estimation, this project has achieved a remarkable level of precision in recognizing and categorizing different food items. Through the use of deep learning algorithms, the system can swiftly analyze input images on the web framework to pinpoint the specific food item within seconds. Additionally, the system can estimate the calorie content of the identified food, equipping users with essential information to effectively monitor their dietary intake. The intelligent diet monitoring capabilities of SmartBite empower users to make well-informed decisions regarding their food selections. By monitoring and evaluating their daily food consumption, users can gain valuable insights into their nutritional patterns, establish personalized objectives, and make necessary adjustments to attain a well-rounded and healthy diet.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Early Detection, Deep learning,CNN Algorithm, Image Processing.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

We propose a computer vision-driven approach aimed at automating the detection of cracks on railway tracks to enhance inspection and security measures.Beginning with the acquisition of images using digital cameras, the system employs pre-processing methods like color transformation and noise removal to enhance image quality. Image segmentation isolates crucial regions using techniques such as Canny edge detection, while feature extraction utilizes advanced models like ResNet and Darknet to capture intricate patterns. Deep learning algorithms, including YOLOv5 and CNN, facilitate real-time object detection and classification, with YOLO focusing on high-probability areas and CNN performing classification based on extracted features. The proposed algorithm demonstrates a detection accuracy of 94.9% on the acquired images, with an overarching error rate of only 1.5%.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

YOLO, ResNet , DarkNet ,CNN

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Cyberbullying is a severe problem that impacts teens and adults on the internet.It has led to incidents like sadness and suicide.The demand for social media platform content regulation is expanding. To develop a model based on the use of natural language processing to identify cyberbullying in text data and machine learning, the following study uses data from two different types of cyberbullying: hate speech tweets from Twittter and comments based on personal assaults from Wikipedia forums. To determine the optimal strategy, three feature extraction techniques and four classifiers are examined. The model yields accuracy levels over 90%.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Wikipedia, Twitter, machine learning, hate speech, personal attacks, and cyberbullying

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The most common kind of cancer in women, breast cancer is bad for both physical and emotional health in those who have it.. Breast cancer remains a significant health concern worldwide, with early detection crucial for successful treatment. Due to complexities present in Breast Cancer images, image processing technique is required for detecting cancer. New deep learning techniques were needed for early breast cancer detection. Histopathological pictures are used as the dataset for this research. Breast tissue histopathological investigation is essential for the diagnosis of breast cancer. This project aims to develop a web application using Django, a Python-based web framework, for managing and viewing breast cancer histopathological images. Images are processed using histogram normalization techniques. This project implements the Convolutional Neural Network (CNN) model based on deep learning and helps in improving the efficiency of breast cancer diagnosis.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Early Detection, Deep learning, Histopathological Images, CNN Algorithm, Image Processing.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

An blood group measurement method is in high demand worldwide, with developing nations having the greatest need for this kind of technology. Building this solution would be a good use for image processing, which is the most widely used device in both resource-rich and resource-poor places. A noninvasive method of measuring blood group is proposed in this project. In order to determine blood groups, it is also compared how different data collection locations,CNN,biosignal processing methods, theoretical underpinnings, photoplethysmogram (PPG) signal and features extraction procedures, image processing algorithms, and detection models vary. The results of this research were then utilized to suggest practical strategies for developing a noninvasive point-of-care tool for blood group assessment based on image processing.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Non invasive, biosignal processing, photoplethysmogram ,Image Processing.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Machine learning has application in many different fields worldwide. It is crucial for healthcare professionals to utilize machine learning algorithms and data analysis tools in order to improve patient outcomes and provide more accurate diagnoses. Such knowledge, if anticipated well in advance, can provide physicians with vital intuitions, allowing them to adjust their diagnosis and method specific to each patient. Using machine learning techniques, we are attempting to predict potential cardiac disorders in humans. This study compares the effectiveness of variety of classifiers, comprising of the Random Forest, SVM, KNN, and logistic regression. We also provide an ensemble classifier that combines the best features of both strong and weak identified Modes to conduct hybrid classification, as it may need an abundance of features.

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Machine learning, Heart failure.

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

The internet's explosive growth has resulted in a rise in cyberthreats, with phishing assaults presenting a serious risk to both people and businesses. Phishing websites aim to trick visitors into disclosing private information, including bank account information and login credentials. Real-time detection of these harmful websites is essential to protecting consumers' privacy and security when they are online. Online security is seriously threatened by the frequency of phishing attempts, in which malevolent actors try to obtain personal information by pretending to be reputable websites. In order to counter this threat, this project presents an extensible and open-source system that uses an artificial neural network (ANN) to detect phishing websites. The goal of the phishing website detection system is to accurately distinguish between legitimate and phishing websites, improving the capacity to safeguard internet users from malicious attacks

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

PHISHING WEBSITE DETECTION USING DEEP LEARNING

  License

Creative Commons Attribution 4.0 and The Open Definition

 Abstract

Stress, which is a more and more common part of contemporary life, can have a serious negative effect on a person's physical and mental health. Determining and tracking stress levels is therefore essential to improving general health and quality of life. The "Human Stress Detection Based on Sleeping Habits Using Machine Learning with Random Forest Classifier" project offers a cutting-edge and successful method for determining a person's degree of stress by looking at how they sleep. Utilizing the robust features of the Python programming language, the research makes use of the Random Forest Classifier algorithm, which is renowned for its adaptability and precision in classification assignments .This project's primary objective is to develop a reliable stress detection system 4 that can provide insightful data about people's stress levels, enabling timely interventions and promoting improved mental health. Numerous significant variables related to stress levels and sleep patterns are included in the dataset that was carefully chosen for the study. The user's snoring range, respiration rate, body temperature, limb movement rate, blood oxygen levels, eye movement, heart rate, number of hours slept, and stress levels--which are divided into five classes--are among these parameters. The classes are 0 (low/normal), 1 (medium low), 2 (medium), 3 (medium high), and 4 (high). By including these several criteria, a thorough examination of sleep patterns and their relationship to stress levels is ensured. The model was able to learn complex patterns from the dataset and forecast stress accurately based on the user's sleeping patterns, as seen by the high accuracy that was attain. Research and treatments in medicine as well as personal health monitoring are just a few of the many possible uses for this stress detection system. People can take proactive steps to reduce stress, enhance sleep quality, and promote general well-being by using the system to analyze their sleep patterns and receive insights into their stress levels

Licence: creative commons attribution 4.0

  License

Creative Commons Attribution 4.0 and The Open Definition

 Keywords

Random Forest Classifier algorithm ,Decision Tree

  License

Creative Commons Attribution 4.0 and The Open Definition