Guideline for Annotating Dentofacial Markers to Distinguish Real and Synthetic Faces
Year: 2025
Type: article
Abstract: This document serves as the official annotation guideline for labeling dentofacial attributes in facial images, with the objective of distinguishing between real human portraits and AI-generated faces. It was developed as supplementary material for the research article "Dentofacial Markers for the Detection of Faces Generated by Diffusion-Based Text-to-Image Artificial Intelligence Models." The guideline defines a standardized set of 12 dentofacial markers—including tooth shape, eye gaze consistency, pupil symmetry, and skin pore visibility—and provides operational definitions, applicability criteria, and examples to ensure consistent and reproducible annotations. This resource is intended for use by researchers and annotators working on computational image analysis, deepfake detection, and medical forensics involving facial imagery.
Authors: Villena, Fabián, Calizaya, Lilian, Rossi, Brunella, Dunstan, Jocelyn, Aguayo, Sebastián
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6949224321
FWCI:
Cites:
Cited by:
Related to: