Deep Learning


Real-time Market Analysis with Online Learning

Introduction The financial world is characterized by ever-changing market dynamics. Predicting market exchange rates in real-time is a challenging yet essential task for investors, traders, and financial institutions. This project leverages online learning, particularly the SGDRegressor model, to make real-time predictions of market exchange rates, bridging the gap between finance and technology. The Challenge Volatile […]

Real-time Market Analysis with Online Learning Read More »


Arabic CRNN based OCR

Introduction The Arabic CRNN-based OCR project introduces an innovative approach to Optical Character Recognition (OCR) for the Arabic language. By utilizing a deep Convolutional Recurrent Neural Network (CRNN) coupled with Connectionist Temporal Classification (CTC) loss, this project sets a new benchmark for recognizing and segmenting Arabic text within images, a task notably challenging due to

Arabic CRNN based OCR Read More »


Social Event Detection using GNN

Introduction The project of Social Event Detection using GNN represents a sophisticated amalgamation of deep learning techniques aimed at incrementally detecting social events via Heterogeneous Graph Neural Networks (GNNs). This initiative, leveraging the KPGNN model with BERT/LSTM embeddings and Graph Isomorphism Convolution (GInConv) layers, strives to revolutionize the way we understand and respond to online

Social Event Detection using GNN Read More »


GAN Image-to-Image Translation

Introduction The GAN Image-to-Image Translation project is an imaginative fusion of technology and art, aimed at transforming grayscale images into lifelike full-color RGB images. Using Generative Adversarial Networks (GAN), this project adds vibrancy and reality to black-and-white images. This case study explores the challenge, the solution devised, and the broad implications of this creative initiative.

GAN Image-to-Image Translation Read More »


TrOCR Fine-tuning for Arabic

Introduction The complexity of Arabic text recognition, especially with diacritics, represents a major challenge in the field of Natural Language Processing (NLP). The TrOCR Fine-tuning for Arabic project takes on this formidable task by employing advanced deep learning models to recognize Arabic text with diacritics. This case study explores the challenges, the innovative solutions, and

TrOCR Fine-tuning for Arabic Read More »


Logo Generator Stable Diffusion

Introduction The field of graphic design has seen the rise of automation and artificial intelligence, and the Logo Generator Stable Diffusion project represents a cutting-edge example of this evolution. The goal of this project is to employ deep learning models to autonomously generate unique, visually appealing, and market-appropriate logos. This case study highlights the challenges

Logo Generator Stable Diffusion Read More »

Scroll to Top