AI/ML Engineer focused on computer vision, reinforcement learning, and evolutionary algorithms. I build robust ML systems and explore the intersection of classical optimization with modern deep learning — from edge inference to multi-agent control.
Research interests: Neural Architecture Search · Distributed Training · Edge AI Optimization
| Cpp-Object-Detection-Yolov5-OpenCV | Real-time YOLOv5 inference engine in C++ via ONNX Runtime — 60+ FPS on edge devices |
| G_ESRGAN | Enhanced ESRGAN for 4× image & video super resolution with perceptual loss tuning |
| Military-Vehicles-Detection | YOLOv5 fine-tuned on a custom military vehicle dataset |
| Breast-Tissue-Cropper-Tools | Medical imaging pipeline for automated mammogram tissue segmentation |
Full collection → RsGoksel/Reinforcement-Learning
| Mechopter | PyGame-based quadcopter simulator with PPO-trained flight controller |
| Reinforcement-Learning-PongGame | PPO agent trained to play Pong from raw frames |
| Snake-Game_PPO-Solution | OpenAI Gym Snake environment solved with PPO — training curves included |
Full collection → RsGoksel/Genetic-Algorithms
| SnakeGame-with-GeneticAlgorithm | Neural network controller for Snake evolved entirely via genetic algorithm |
| Partical-Swarm-Optimisation-Examples | PSO framework with adaptive parameters and convergence analysis |
| Data-Structures-Cpp | Foundational data structures in C++ with complexity analysis and memory profiling |
| HeisenMech_TDD_2024 | Hackathon project — #Acıkhack2024TDDİ |
| Article |
|---|
| Spike Neural Networks — Neuromorphic computing, temporal coding, hardware acceleration |
| Particle Swarm Optimization — Theory, convergence proofs, multi-objective cases |
| Signal Processing for ML — Classical DSP concepts bridged to deep learning |
Open to collaboration on AI research and industrial applications