An interactive, WebGL-accelerated platform for visualizing and experimenting with neural networks in the browser. Built with TensorFlow.js, React, and D3.js.
High-performance WebGL 2.0 visualization framework for interactive 3D fractals and particle-based cosmic simulations, powered by GLSL shaders.
Modular framework for enhancing factual consistency in abstractive summarization using reinforcement learning, post-hoc correction, and factuality-guided decoding.
End-to-end AI platform for drone-based urban waste detection using synthetic data generation, YOLOv8 detection, and interactive analytics. Built on AirSim and Unreal Engine.
Framework enhancing visual odometry and SLAM robustness in drone navigation through simulated urban canyon environments. Combines classical SLAM with deep learning features.
High-performance C++20-based library for training Gradient Boosting Decision Trees from scratch. Features histogram-based split optimization, SIMD acceleration, and multi-threading.
Framework for detecting hallucinations in LLM chain-of-thought reasoning. Features synthetic data corruption, transformer-based classifiers, Streamlit UI, and FastAPI backend.
C++20 framework for training autonomous drone flight controllers using deep reinforcement learning. Combines PPO optimization with high-performance gRPC bridge to Microsoft AirSim.
C++ platform for real-time network traffic monitoring and anomaly detection. Leverages efficient data handling and pattern analysis to identify suspicious behaviors.
State-of-the-art personal health assistant powered by AI, designed to analyze symptoms, identify potential medical conditions, and provide personalized health insights.
Modular and extensible framework supporting real-time face detection, facial landmark localization, and identity recognition through an API-driven architecture.
Real-time intelligent surveillance system designed to detect human falls and anomalous behaviors using deep learning techniques. Includes automated alerting and monitoring dashboard.
Machine learning-powered energy consumption prediction system that analyzes historical data to forecast future energy usage trends, optimizing efficiency and sustainability.
Interactive AI dashboard for machine learning model analysis and explainability. Supports model training, dataset exploration, feature importance analysis, and SHAP-based explanations.
C++17 implementation of a Naive Bayes / Bag-of-Words sentiment analysis classifier pipeline, built with CMake.
Real-time emotion detection system using computer vision and deep learning techniques to analyze facial expressions and classify emotional states.
Face-recognition system that works on images. First of all you need to register valid image as recognition. Built with Streamlit for easy UI interaction.
Teknofest 2024 Application Project - A comprehensive project developed for Turkey's premier technology competition, focusing on innovative solutions and cutting-edge development.
Website that I designed for Google DSC while I was a club leader. Features modern design with JavaScript animations and interactive canvas elements.
CPExcel is a Windows application to modify Excel sheets with a user-friendly interface built using PyQt5. Provides efficient spreadsheet manipulation capabilities.
UTAB Uni is a website where professors in university access their assigned answer sheets of students, do modifications, grade results and download as XLS format.
Machine learning project for HR analytics and employee prediction. Uses advanced statistical models to predict employee behavior and performance metrics.