YOLOv12 Object Detection Model — Reproducible Training, Evaluation & Results Repository

This repository contains the complete experimental package used to train, evaluate and analyze a YOLOv12 object detection model. It includes dataset partitions (train/validation/test), model configuration files, pre-trained base weights, the final optimized model checkpoint, training logs, loss metrics, and a fully reproducible Jupyter notebook.

All material enables complete reconstruction of the training process, from dataset preparation to evaluation over the held-out test set, ensuring transparency and experimental reproducibility. The included results.csv file allows plotting loss evolution curves over epochs, while opt_results.csv documents the hyperparameter optimization stage. The repository also stores YOLO-generated performance outputs inside the data/ folder, supporting qualitative inspection of model behaviour.

This resource is intended for research, benchmarking, educational use, and extension into new computer vision applications. A Bayesian optimization module will be added for advanced hyperparameter search experiments.