YOLO11-EMME v2.0: A 10,000-image Eye-Tracking Dataset for Archaeology, Ophthalmology and HCI

This dataset provides high-resolution eye-tracking images with iris–pupil annotations and pilot gaze metadata (n=27). It includes split files for training/validation under varied lighting/occlusion conditions. Reproducible code and figures correspond to the linked GitHub repository. • Format: YOLO (class_id x_center y_center width height), normalized [0,1]• Classes: iris (0), pupil (1)• Folder structure: images/{train,val}, labels/{train,val}, sample_images/• Reproduction: pip install -r requirements.txtpython run_inference.py --weights results/weights/best.pt \ --source datasets/sample_images --imgsz 640 --save --seed 42 Outputs: runs/predict/*• Files included: results/figures/*.png, datasets/participants.csv, README_DATASET.md, results_summary.csv, best.pt• License: MIT• FAIR: public access, versioned DOI, code on GitHub GitHub (software): https://github.com/seungjin-data/Advancing-Eye-Analysis-and-Eye-Tracking-with-Ultralytics-YOLO11How to cite (this version): 10.5281/zenodo.17310457All-versions DOI: 10.5281/zenodo.17120642 Seung Jin Kim — ORCID: 0009-0007-6876-0777 — Affiliation: AI Convergence Engineering, Assist University, Seoul, Republic of Korea — Role: Author/Contact Myuhng Joo Kim — Affiliation: Department of Computer Science, Assist University, Seoul, Republic of Korea — Role: Author