Subject: Feature Suggestion: Forensic Reconstruction Mode for Historical and Archaeological Images
Dear OpenAI Team,
I would like to submit a feature suggestion regarding image generation and reconstruction tools.
When reconstructing historical figures from sculptures (for example Roman busts), current models tend to idealize facial features. Generated faces often appear more symmetrical, younger, or more conventionally attractive than the original sculpture suggests.
For historians, archaeologists, and researchers this is a significant limitation, because Roman portrait sculpture (verism) intentionally preserved imperfections, asymmetries, and realistic facial structures.
A potentially valuable feature would be a “Forensic Reconstruction Mode” designed specifically for scientific and historical applications.
Such a mode could include:
• suppression of automatic beautification or aesthetic smoothing
• preservation of asymmetries present in the source sculpture
• geometric constraints derived directly from the sculpture’s proportions
• optional anthropological parameters (e.g., age range, body-fat estimation, regional phenotype statistics)
• generation of multiple statistically plausible variants rather than a single idealized face
This type of functionality could have applications in:
• archaeology and museums
• historical documentaries and media production
• academic research
• forensic science
• educational visualization of historical figures
From a product perspective, such a tool could potentially become a specialized professional feature or API offering for museums, researchers, and media production teams.
I am a photographer working with historical imagery and Roman sculpture, and I frequently encounter this limitation when attempting realistic reconstructions.
Thank you for considering this suggestion.
Best regards,
Jörg Sänger
Germany