Project Details

AI Projects

Artificial Intelligence solutions and applications

Project Information

View on GitHub
  • Category: AI/Computer Vision
  • Technologies: OpenCV PyTorch Python
  • Project Date: April 2024
  • Dataset: OPA Dataset

Advanced Object Placement Model

This project implements an advanced object placement system that automatically places an object into a background image with realistic positioning and lighting.

Technical Details

  • Image preprocessing and segmentation
  • Trained on OPA dataset with 50,000+ images
  • Shadow generation based on light source estimation
  • Multi-threaded CPU and GPU compatible implementation

Key Features

  • Automatic object segmentation and extraction
  • Physics-based placement suggestions
  • Real-time preview of placement results
  • Batch processing for multiple objects
Object Placement Interface
Object Placement Example

Machine Learning Projects

Predictive modeling and data analysis solutions

Project Information

View on GitHub
  • Category: Machine Learning
  • Technologies: Scikit-learn Numpy Pandas Flask
  • Project Date: February 2024
  • Dataset: IMDB Reviews
  • Accuracy: 89.2% on test set

Movie Review Sentiment Analysis

A machine learning model that classifies movie reviews as positive or negative with high accuracy.

Technical Approach

  • Trained on 50,000 reviews dataset
  • Custom word embeddings trained on movie-specific corpus
  • Implemented using TF-IDF and Logistic Regression
  • Deployed as a Flask web application

Key Features

  • Real-time sentiment prediction API
  • Interactive visualization of sentiment trends
  • Batch processing for large datasets
  • Model performance metrics dashboard
Sentiment Analysis Interface
Model Performance

Project Information

View on GitHub
  • Category: Healthcare ML
  • Technologies: Scikit-learn Pandas Numpy XGBoost
  • Project Date: January 2024
  • Dataset: UCI Heart Disease Dataset
  • Accuracy: 87.5% (5-fold cross-validation)

Heart Disease Predictor

A machine learning model that predicts the likelihood of heart disease based on medical attributes and lifestyle factors.

Technical Implementation

  • Evaluated 8 different classification algorithms
  • Comprehensive feature engineering pipeline
  • Risk probability calibration system
  • SHAP values for explainable AI

Key Features

  • Interactive feature importance visualization
  • Individualized risk assessment reports
  • Exportable prediction results
  • Responsive web interface for medical professionals
Heart Disease Predictor Interface
Feature Importance

Deep Learning Projects

Neural networks and computer vision solutions

Project Information

View on GitHub
  • Category: Deep Learning
  • Technologies: TensorFlow Transfer Learning Keras
  • Project Date: March 2024
  • Dataset: Kaggle Dog Vision

Dog Breed Identification Project

A sophisticated deep learning model capable of identifying 120 different dog breeds from images with high accuracy.

Technical Details

  • MobileNetV2 base model with custom classification layers
  • 92% accuracy on validation set
  • Data augmentation techniques for improved generalization
  • User-friendly interface for real-time predictions

Key Features

  • Real-time breed prediction from images
  • Top 3 predictions with confidence scores
  • Breed information database integration
  • Mobile-friendly responsive design
Dog Breed Identification Interface
Dog vision

Project Information

View on GitHub
  • Category: Deep Learning/Computer Vision
  • Technologies: OpenCV FaceNet Python
  • Project Date: December 2023
  • Dataset: Custom Collected (50 identities)
  • Accuracy: 94.3% on test set

Face Detection and Recognition System

A face recognition system that detects faces and matches them against a database of known individuals.

Technical Architecture

  • Two-stage pipeline: detection → recognition
  • Haar cascades for face detection
  • FaceNet embeddings for face representation
  • Cosine similarity for matching

Key Features

  • Real-time face detection and recognition
  • Database management for known individuals
  • Attendance tracking system integration
  • Access control system prototype
Face Detection Interface
Recognition Samples

Web Development Projects

Interactive web applications and platforms

Project Information

View on GitHub
  • Category: Web Development
  • Technologies: HTML5 CSS JavaScript
  • Project Date: May 2025

Pet Friendly City Platform

A web platform that helps pet owners find pet-friendly locations and services in their city.

Technical Implementation

  • ChatBot Integration
  • User Form
  • Upcoming events schedules
  • Adopt A Pet

Key Features

  • Searchable directory of pet-friendly businesses
  • User reviews and ratings system
  • Community events calendar
Pet Friendly City Dashboard
Park Analysis