This project focused on creating software to verify the accuracy of handwritten mathematical notations using Artificial Intelligence and Computer Algebra System. Utilizing Optical Character Recognition technology and Computer Algebra System, the software can recognize, digitize, and check mathematical expressions for errors. It was implemented using Python and multiple technologies, such as Sympy and Tesseract Optical Character Recognition, offering broad applications …
Synthesising Classical Guitar Pieces Using Transformer
After a literature review, no specific AI model for classical guitar in symbolic music production was found. Therefore, a new AI model was developed to generate new pieces while preserving the melodic and harmonic structure based on input notes. To do this, a dataset containing solo classical guitar pieces was tokenized using REMI and augmented. The Music Transformer deep learning …
Protect Art and Creativity: A Prevention Framework for Unauthorized Learning of Text to Image AIs.
In this work, we aim to solve the problem of unauthorized learning of works arising from mass data collection from Text to Image AI models, specifically Stable Diffusion. The TTI model performs indiscriminate web data crawling to collect images, and these images are used for model learning without the consent of the author. The TTI model is capable of learning …
Thermography-based driver assistance system
We have developed a software and device that can reliably detect and alert drivers of potential hazards, such as animals or people on the road or roadside, beyond braking distance, even under poor visibility conditions. We have achieved this through an external device equipped with a thermal camera, which operates at high speed. This information is then transmitted and processed …
An application for the prevention of road accidents
The object of the Project is the construction of an Android mobile application, where traffic accidents will be recorded easily by the police via GPS to inform the citizens in real time. The purpose is to mark the dangerous points of the road network, where repeated accidents occur (Black Spots), so citizens can be warned in time to avoid further …
Web Application Firewall Based on Machine Learning and Artificial Intelligence
This project introduces an advanced web application firewall that combines machine learning, artificial intelligence, and advanced encryption algorithms to provide real-time protection against cyber-attacks. It offers automated self-penetration testing, mutation-based threat detection, and a unique cryptography technique for data security. The firewall’s multifunctional approach and cost-effectiveness contribute to a safer digital environment for individuals, businesses, and organizations. Andria Kelekhsaevi Alexandre …
Online student calendar – iSCAL.sk
Attendance in competitions, lectures and open-door days significantly helps students with their individual growth and knowledge gains. There are many activities offered by institutions, which is certainly a positive fact. However, it simultaneously causes trouble for people to find important information such as date, venue or form the event will be held in. Each event has its own landing page, …
MegaView
MegaView is an offline app (without an internet or database connection) for mobile phones, tablets, fitness bracelets, smartwatches, or any other electronic device that can transmit audio messages providing specific information to visually impaired individuals. The app can be easily customized based on usage (wardrobe, traffic circulation, books, etc.). By utilizing real-time object recognition technology (YOLO and TensorFlow Lite, OCR), …
Future of Autism
With a prevalence of 1 per 100. Autism represents a global challenge. Behavioral improvement with specialists is the only proven method with high socioeconomic challenges. Therefore, an application was created that assesses the condition of a case with autism. It uses GARS, followed by filling Stanford-Binet Intelligence Scale 5th edition report. It makes a custom program suitable for the case …
A Novel Method of Constructing Block Ciphers Resistant to Linear and Differential Cryptanalysis
I have pioneered a groundbreaking method for creating ciphers that adeptly manage the encryption of large data chunks, an issue that was previously quite challenging. In the past, we faced a trade-off between efficiency and security when it came to ciphers with large block sizes. This cutting-edge approach not only offers faster performance, but is also mathematically proven to be …
- Page 1 of 2
- 1
- 2