International Journal of Computer Engineering and Applications
International Journal of Computer Engineering and Applications
IMPLEMENTATION OF CONVERTING UNICODE TEXT TO ORIGINAL TEXT
Mr. Prashanth Kumar HM¹ & Dr. Subramanya Bhat S²
¹Research Scholar, College of Computer Science, Srinivas University, Mangalore, India
²Professor, College of Computer Science, Srinivas University, Mangalore, India
Keywords - Unicode, Jumbled Text, Conversion, ASCII, Encoding, Interface, Index.
ABSTRACT
In regional plagiarism checker initially, we must understand for character conversion, almost we have 3000+ regional conversion font available in market. More fonts may refer to Unicode character, UTF or encoding format. Other font won’t create any abjection while converting but we may more concentrate on Unicode. Because most regional language text creation is under Unicode format. Unicode is an international standard for the representation of text and characters in computers and other devices. It maps characters to unique numerical values, called code points, allowing computers to store, process, and display text in a consistent way, regardless of the language or script in which it was written. Unicode supports a wide variety of languages, scripts, symbols, and special characters, including emoji, mathematical symbols, and bidirectional text. It also supports character encoding, allowing characters to be stored, processed, and displayed in a way that is compatible with different platforms and devices. In our implementation part Unicode conversion is an essential tool for text processing, storage, and display, making it easier and more consistent for the need to use text in different languages. Finally, it’s challenging work for converting Unicode text to original format into the next level of plagiarism process.
VIRTUAL FITNESS TRAINER
Prof. Shrishail Patil¹, Rushikesh Cheke², Rushikesh Mane³, Krushna Godse4, Arti Bangar5, Vrushabh Malwade6
Student, Assistant Professor, Dept. of Computer Engineering, JSPM’s Bhivarabai Sawant Institute Of Technology & Research Wagholi, Maharashtra, India
Keywords - Trainer, AI, Machine Learning, Deep Learning, Fitness.
ABSTRACT
The fourth industrial revolution and the ongoing creation of new technology have made sedentary existence possible. As a result, non communicable diseases like diabetes, cancer, and cardiovascular and respiratory disorders are now reaching epidemic levels. Clients are screened by a traditional gym teacher who then recommends training regimens that can help them reduce their risk of non communicable lifestyle diseases. Unfortunately, fitness instructors are frequently expensive and not always accessible, available, or inexpensive. This study looked into whether a method for creating a more affordable workout regimen might be developed using today's computing power. Virtual Fitness Trainer. Up to four metres away from the camera, the system displayed flawless real-time object detection and tracking, and it also generated results for a distance of up to eight metres.
POVERTY ANALYSIS, PREDICTION USING MACHINE LEARNING METHODS
Prof. V. D. Ghonge, Parth Sandeep Kadam, Vitthal Arjun Bhakare, Amit Kailas Bodake, Prashant Sudhakar Warungase
Department of information technology, Smt. Kashibai Navale College of Engineering. Maharashtra, Pune -411041, India
Keywords - Poverty Prediction, Machine Learning, Algorithms, datasets, Prediction.
ABSTRACT
Lack of sufficient resources to provide for basic needs like food, clean water, housing, and clothing, as well as in today's world, access to health care, education, and even transportation, is referred to as poverty. The government of the country was given many methods, but they don't function as they should. The predictions are inaccurate, and the country's customary method of making predictions involves conducting a site survey, which is pricy and labor intensive and sometimes a waste of time before learning the actual outcome. Making educated policy decisions and efficiently distributing resources to the places that need assistance the most is severely hampered by the absence of credible statistics on poverty in the nation. In order to fully understand the causes of poverty, we will first conduct a multidimensional analysis of poverty using multiple correspondence analysis. Next, we will make predictions using three different machine learning techniques, and finally, based on prior research, we will also use satellite images processed through convolutional neural networks to estimate the level of poverty. In order to determine whether the method is better suited to comprehend and predict poverty in a country, this paper compares simple machine learning methods to advanced deep learning methods in an effort to build on prior research.
TRACKING THE IMPACT OF PM POSHAN ON CHILD’S HEALTH
Prof. Megha R. Mehar, Arya pathrikar, Rohan Chavan, Pruthvi Inamke, Siddarth Koul
Department of information technology, International Institute of Information Technology. Maharashtra, Pune-411057, India
Keywords - Deep Learning, R-CNN, Machine learning, OpenCV, Algorithms.
ABSTRACT
Using our model to keep track of the daily protein, calorie, and weekly menu intake for each child. Make sure kids are going to school and getting at least one nutritious meal per day for their physical and mental development. Provide interested authorities access to a streaming service upon request so they may verify the status of the programme and acquire insights by monitoring and tracking the health of the pupils.
The Pradhan Mantri Poshan Shakti Nirman (PM POSHAN) Centrally Sponsored Programme would provide one hot cooked meal per day in Government and Government-aided Schools from 2021–2025. The Ministry of Education is carrying out the Plan. In addition to the 11.80 crore students in classes I through VIII attending the 11.20 lakh schools, the Plan also provides hot prepared meals to children in pre-schools or Bal Vatika (before class I) at primary schools. Without regard to gender or social background, the Plan is applied to all eligible children throughout the nation. The primary goals of the PM POSHAN Scheme (previously known as the Mid-Day Meal Scheme) are to address two of the most pressing issues facing the majority of children in India, namely hunger and education, by enhancing the nutritional status of eligible students in Government and Government-aided schools as well as encouraging low-income students from disadvantaged sections to attend school more frequently and support their ability to concentrate in the classroom. Calorie estimation from food photos using computer vision has recently been developed. However, the volume and mass of foods are not currently recorded in food image datasets, which results in an inaccurate calorie estimation. Using a deep learning algorithm for food detection and a novel food image dataset containing volume and mass records of foods, we provide in this study a full calorie estimation. Every image in our dataset has related food annotation, volume, and mass records as well as a specific calibration reference. Our collection consists of different photographs. In order to calculate the number of calories in the proposed dataset, we propose a novel food image dataset including records of food's volume and mass in this study. A deep learning technique employing Faster R-CNN is utilized to detect the meal and calibration item in the proposed dataset, after which we estimate each food's volume and calories. The outcomes of the experiment demonstrate the efficacy of our estimation technique. This dataset is the first publicly available set of food image data that can be used to assess calorie calculation techniques based on computer vision.
BLOCK CHAIN-BASED CROWDFUNDING SOLUTION
Prof. Keshav Tambre, Akhil Bhalerao, Aseem Khandekar, Radhika Bhutra, Varshitha
Department of information technology, International Institute of Information Technology. Maharashtra, Pune, India
Keywords - Crowdfunding, Block chain Technology, Synergy, Smart Contracts, NFT.
ABSTRACT
The utilization of Block chain technology is on the rise in the 21st century as an underlying technology for Information Communication and Computation (ICCT)[1], and it has the potential to benefit various industry sectors, including primary, secondary, tertiary, and quaternary industries. This study has focused on exploring specific potential applications of Block chain technology principles in the finance industry. This paper discusses the potential of using smart contracts to automate the execution of crowdfunding transactions and to ensure that investor funds are released only after the crowdfunding goal is met. Block chain has applications in a variety of industries, including the energy sector, forestry, fishing, mining, material recycling, air pollution monitoring, supply chain management, and their related operations. Block chain is essentially a trusted and decentralized database. We give a survey of Block chain-based network applications in this study. Our objective is to cover the development of Block chain-based solutions that aim to revitalize the current, largely centralized, network application area. When we use Block chain to reimagine the area, we outline a number of typical difficulties, traps, and shortfalls that could arise. We want this to serve as a helpful reference tool for anyone considering switching to a Block chain-based solution for their current use case or building an automated one from scratch. Finally, the paper will focus on existing and upcoming developments in this area, and consider the potential future of Block chain-based crowdfunding solutions.
IOT BASED WALKING STICK FOR VISUALLY IMPARED USING RASPBERRY PI
Prof. S.A. Itkarkar, Revati Gajbhar, Aishwarya Hatkar, Gauri Jadhao
Department of Electronics & Telecommunication, Bharati vidyapeeth’s college engineering for women, Pune, India
Keywords - Raspberry Pi, Ultrasonic sensor, GPS Module, Earphone.
ABSTRACT
A Smart stick system concept is devised to provide a smart electronic aid for blind people. Blind and visually impaired people find difficulties in detecting obstacles during walking in the street. The system is intended to provide artificial vision and object detection, real time assistance via GPS by making use of Raspberry Pi. The system consists of ultrasonic sensors, GPS module, and the feedback is receive through audio, voice output works through TTS (text to speech). The proposed system detects an object around them and sends feedback in the form of speech, warning messages via earphone and also provides navigation to specific location through GPS. The aim of the overall system is to provide a low cost and efficient navigation and obstacle detection aid for blind which gives a sense of artificial vision by providing information about the environmental scenario of static and dynamic object around them, so that they can walk independently.
PUBLIC GARDEN AUTOMATION
Prof. S. G. Nagpure, Shalaka Shrikant Khalate, Poonam Santosh Shinde, Shweta Hanumant Malpote
Department of Electronics & Telecommunication, DY Patil School of Engineering Academy, Ambi Pune, India
Keywords - Microcontroller & LED, Keyboard, Relay.
ABSTRACT
This project the idea of Public Garden Automation technology has been presented. Avoid wastage of water and electricity. This project is a fine combination of analog and digital electronics. As a part of our fourth year circular activity we are making the project whose title is “Public Garden Automation”. We have used Microcontroller as a main component of the project. Now a Microcontroller has become a main component of many of the electronic circuits. Also Liquid Crystal Display (LCD) is used on major basis for the display purpose. . We are using time for each and every parameter so that only at that instant it can resume again. For controlling this technique we are using a microcontroller which will be going to control all the parameters. For light we are using LDR sensor, for water level sensor and for particular control we are using various pumps and the motors.
DYNAMIC TRAFFIC SIGNALING SYSTEM
Dr. D. S. Mantri¹, Saurabh Khamkar², Sharau Moon³
Professor, Department of Electronics and Technology1,
Student, Department of Electronics and Technology2,3
Sinhgad Institute of Technology, Lonavala, Maharashtra, India
Keywords - Traffic control, Raspberry Pi, Camera, Traffic light.
ABSTRACT
The most significant issue which is being looked at by the advanced world is the traffic blockage in the City communities and towns. The system attempts to reduce the possibility of traffic jams, caused by traffic lights, to a limit. The system is contingent on count of vehicles and comprises of raspberry-pi which positively analyses the situation as a result of which the traffic lights delay is altered for each lane. Thus, it updates distinguish ranges for traffic light delays and sets those accordingly. The cameras are placed at traffic intersections for analyzing the traffic thickness. A camera is placed along with the traffic light. It catches picture groupings. Picture handling is a pompous technique to tackle the transition of the traffic light. In presence of an emergency vehicle the red sign on the traffic signal turns green with assistance of Message Queuing Telemetry Transport (MQTT) which gives a reasonable method of convenience to crisis vehicles. This recorded vehicle count data can be used in future also to investigate traffic conditions at respective traffic lights connected to the system. For germane analysis, the record data can be downloaded via interaction between the computer and the raspberry-pi after which it will send the appropriate signal to the LED light system. In the future this technique can be often used to enlighten individuals about traffic conditions at different locations.
DESIGNING DISEASE PREDICTION MODEL USING MACHINE LEARNING APPROACH
Prof. Sarika Aundhakar, Lokesh Girase, Rishikesh Khakal, Vaishnavi Wattamwar
Department of Computer engineering, Smt. Kashibai Navale College of Engineering. Maharashtra, Pune
Keywords - RANDOM FOREST, NAIVE BAYES, DECISION TREE, Machine learning, Disease Prediction.
ABSTRACT
Today, human beings face diverse sicknesses because of the environmental situation and their residing conduct. So the prediction of sickness at an in advanced degree will became a vital task. But the correct prediction on the idea of signs will become too hard for doctors. The accurate prediction of sickness is the maximum hard task. To triumph over this trouble, information mining performs a vital function to are expecting the sickness. Medical technological know-how has a massive quantity of information increase according to year. Due to the multiplied quantity of information increase with inside the clinical and healthcare area the correct evaluation of clinical information has been cashing in on early affected person care. With the assistance of sickness information, information mining reveals hidden sample facts in a large quantity of clinical information. We proposed fashionable sickness prediction primarily based totally at the signs of the affected person. For the sickness prediction, we use Naive bayes and Random Forest gadget studying set of rules for the correct prediction of sickness. For sickness prediction required sickness signs dataset. In this fashionable sickness prediction, the residing conduct of someone and checkup facts don't forget for the correct prediction. The accuracy of fashionable sickness prediction via way of means of the use of RANDOM FOREST is 84.5% that is extra than the NAIVE BAYES set of rules. And the time and the reminiscence requirement also are extra in NAIVE BAYES than RANDOM FOREST, DECISION TREE. After fashionable sickness prediction, this device is ready t deliver the danger related to a fashionable sickness that is a decrease danger of fashionable sickness or higher.
DETECTING LIPS MOVEMENT AND PREDICTING PHRASES USING CNN
Shivanand Gadgi1, Shubhangi Wartale2, Manisha Mane3, Prajakta Narole4 and Vilas Ghonge5
Department of Information Technology, Savitribai Phule Pune University, India
Keywords - Convolutional Neural Network, Deep Learning, Image processing.
ABSTRACT
The audio-visual speech recognition system using lip movement uprooted from side- face images to attempt to increase noise- robustness in mobile surroundings. Although utmost former bimodal speech recognition styles use anterior face (lip) images, these styles aren't easy for druggies since they need to hold a device with a camera in front of their face when talking. Our proposed system landing lip movement using a small camera installed in a handset is more natural, easy and accessible. This system also effectively avoids a drop of signal- to- noise rate (SNR) of input speech. Visual features are uprooted by optic- inflow analysis and combined with audio features in the frame of CNN- grounded recognition.
