Mahcine Learning Developer
My name is Hooman, and I'm a Machine Learning Developer specializing in Deep Learning, Robotics, and Computer Vision. With extensive experience, I have developed innovative solutions in these advanced fields. My projects range from autonomous robotics and intelligent systems to advanced image and video analysis, showcasing my ability to tackle complex challenges. I am dedicated to pushing the boundaries of technology, using state-of-the-art algorithms and frameworks. Explore my portfolio to see the impact of my expertise and the transformative potential of my work.
Learn More!
Thanks to Pytorch and Tensorflow
Hooman's AI Bot
Powered By AI
My Projects
Intelligent Parking Space Management
Intelligent Parking Space Management is a computer vision-based project that detects and manages parking spaces in real-time. Utilizing OpenCV and machine learning, it identifies vacant spots in a parking lot from video footage, making parking management efficient and automated. Ideal for modern smart city solutions
DetectMaster YOLO
Welcome to DetectMaster YOLO, a comprehensive project for object detection using the YOLO (You Only Look Once) algorithm. This repository contains various scripts to detect objects in real-time using a webcam, count cars and people, and detect personal protective equipment (PPE).
Face Detection Model
The Face Detection Model, develops a robust face detection model using advanced machine learning techniques. It involves setting up the environment with TensorFlow, OpenCV, and matplotlib, collecting and augmenting images for a comprehensive dataset, and building and training the model with TensorFlow. The model is fine-tuned for accuracy and evaluated for real-world performance.
Shape Recognition
Shape Recognition is a machine learning project focused on classifying geometric shapes like triangles, stars, squares, and circles. Using TensorFlow and Keras, the project covers the entire pipeline from data preprocessing to model evaluation and prediction, providing visualizations and code snippets to facilitate understanding and implementation.
CO2 Emissions (Linear Regression)
This project focuses on predicting CO2 emissions of vehicles based on their engine sizes using a linear regression model.
Emotion Recognition (CNN)
This project focuses on classifying emotions as happy or sad using machine learning. It leverages TensorFlow and OpenCV to build and train a neural network model. The model processes images of faces, identifying emotional states with high accuracy. The project includes a comprehensive pipeline for data preprocessing, model training, and evaluation, showcasing performance through accuracy and loss curves. This innovative approach aids in understanding and interpreting human emotions computationally.
Binary Handwritten Digit Recognition
This project implements a neural network model using TensorFlow and Keras to recognize handwritten digits as either 0 or 1. The model is trained, evaluated, and tested using the MNIST dataset, demonstrating effective binary classification performance.
Handwritten Digit Recognition
This project implements a machine learning model to recognize handwritten digits from images. Utilizing popular libraries like TensorFlow and Keras, the project demonstrates image preprocessing, model training, and evaluation, providing an end-to-end solution for digit classification tasks.
Softmax Neural Network
The project demonstrates the implementation of the softmax function in a neural network using TensorFlow. It includes data generation, model training, and prediction steps, showcasing the process of building and evaluating a multi-class classification model on synthetic data generated with the `make_blobs` function.
Anomaly Detection in Server Performance Metrics
This project implements an anomaly detection algorithm to identify unusual behavior in server computers using throughput and latency metrics. It aims to enhance server reliability and performance monitoring.
CO2 Emission Multiple Linear Regression
This project involves developing a multiple linear regression model to predict CO2 emissions. Using a dataset of various factors influencing emissions, the model analyzes relationships and quantifies the impact of each predictor, enabling accurate forecasting and informed environmental decision-making.
Coffee Roasting Deeplearning
The Coffee-Roasting-Deeplearning project innovates coffee roasting through advanced deep learning, optimizing temperature and time for superior flavor, aroma, and quality.
CO2 Emissions Linear Regression
This project focuses on predicting CO2 emissions of vehicles based on their engine sizes using a linear regression model.
Cat Recognition (DeepLearning)
The Cat Recognition Project utilizes deep learning techniques to build a neural network capable of recognizing cats in images. Through data preparation, model building, and evaluation, this project demonstrates the application of neural networks in image classification tasks.
Hand Tracking Mediapipe
This project is a real-time hand tracking application using OpenCV and MediaPipe. It captures video from a webcam, detects hand landmarks, and displays the results with FPS. The project includes a main script for video capture and a module for hand detection functionality.
Cat Recognition (Logistic Regression)
This project implements a logistic regression model to recognize cats in images. It involves loading and preprocessing data, initializing parameters, computing the weighted sum plus bias, applying the sigmoid function, and making predictions, with visualizations of cost reduction over iterations.
AI Virtual Mouse
This project leverages computer vision and AI to create a virtual mouse controlled by hand gestures. Utilizing a webcam, the system tracks hand movements, interprets gestures, and performs mouse operations like moving the cursor and clicking, providing a touchless interface for interacting with a computer.