This repository contains the implementation of a project aimed at head pose estimation using a custom Convolutional Neural Network (CNN). The project utilizes the Columbia Gaze dataset, involving steps from data preprocessing, creating synthetic labels for attention and distraction, to training a custom CNN model.
This project aims to estimate head poses (attention and distraction)…
This code performs classification using a Support Vector Machine (SVM) algorithm combined with CURE-SMOTE oversampling.
It is designed to handle imbalanced datasets and evaluate the model performance through k-fold cross-validation. GitHub: https://github.com/aidamohseni/SMOTE-Enhanced-SVM-Classification-with-Stratified-K-Fold
This is a project aimed at recognizing Nastaliq handwritten script, a form of Persian calligraphy. The project involves detecting writing items using contours and regrouping them based on geometric proximity.
It uses a custom dataset of Nastaliq words and character combinations along with a CNN-BDLSTM hybrid model to recognize the text. GitHub: https://github.com/aidamohseni/Handwriting_Nastaliq_Recognition_BLSTM
This project aimed at recognizing Persian handwritten characters. The project employs a hybrid model integrating Convolutional Neural Networks (CNNs) and Bidirectional Long Short-Term Memory (BLSTM) networks to capture both spatial and sequential features essential for accurate recognition of Persian script.
The model is evaluated using accuracy, precision, and recall metrics. The results are validated using…
Using a large enough dataset is so important in a neural network based handwriting recognition solutions. On the other hand, producing natural datasets is so complicated and time-consuming. In this paper, we compare the effect of using synthesized datasets and natural datasets in state of the art neural network based solutions. For natural solution based…
In this paper, we review classic handwriting recognition processes and controversial issues by details. In addition, since there are presented the state of the art approaches based on neural networks solutions, we review RNN approaches. Furthermore, as the database is an important factor in RNN based solutions, we survey the most important datasets for the…
Optical Character Recognition (OCR) technology has undergone a profound transformation, largely thanks to advancements in machine learning. From digitizing printed documents to recognizing text in natural scenes, OCR's scope has expanded, opening new possibilities for data extraction, automation, and accessibility. The Evolution of OCR Technology Historically, OCR systems relied on rule-based algorithms to recognize text,…
In the development of machine learning models, evaluating performance accurately is critical to achieving reliable and robust outcomes. Model evaluation metrics play a fundamental role in this process, providing quantitative ways to measure and compare the performance of models across different tasks and datasets. Understanding Model Evaluation Metrics Model evaluation metrics can be broadly categorized…
As artificial intelligence (AI) systems become more prevalent in critical decision-making processes, the need for transparency and understandability in machine learning models has never been more acute. Explainable AI (XAI) seeks to address this challenge by making the workings of complex models accessible and interpretable to humans, fostering trust and enabling more informed decisions. The…
Transfer learning is a revolutionary approach in machine learning that involves taking a pre-trained model on one task and repurposing it for another related task. This method has significantly impacted the field by reducing the need for large labeled datasets and extensive computational resources, making advanced machine learning more accessible and efficient. Understanding Transfer Learning…
The field of machine learning has experienced exponential growth over the past few decades, evolving from simple linear models to complex neural networks capable of performing tasks that were once deemed impossible for machines. This evolution has been marked by significant milestones, each introducing new capabilities and opening up unprecedented applications across industries. The Beginnings:…
The intersection of quantum computing and artificial intelligence (AI) marks a thrilling frontier in technology. Quantum machine learning (QML) leverages the principles of quantum mechanics to enhance machine learning algorithms, offering a glimpse into a future where computational limits are vastly expanded. This post explores the potential of QML to revolutionize data processing, algorithm efficiency,…