Funding
Funding not disclosed
Founders
Product
Problem
Many significant artworks remain hidden beneath existing paintings, inaccessible to art historians and the public. This loss of cultural heritage limits our understanding of artistic evolution and the full scope of an artist's oeuvre.
Solution
Oxia Palus employs advanced spectroscopic imaging, artificial intelligence, and 3D printing to digitally reconstruct lost masterpieces. This process involves analyzing multispectral data from paintings to identify underlying layers, then utilizing generative adversarial networks (GANs) and neural style transfer to recreate the hidden artwork. The resulting digital reconstructions are then materialized as high-fidelity 3D prints on canvas, effectively resurrecting these dormant pieces. This technology expands the known art historical record by revealing works that would otherwise remain undiscovered.
Target Audience
Primary customers include art museums, galleries, art historians, collectors, and cultural institutions seeking to expand their collections and research capabilities with previously inaccessible artworks.
Features
- Spectroscopic imaging to capture multispectral data from existing artworks.
- Conditional Generative Adversarial Networks (GANs) for image reconstruction and style transfer.
- Neural Style Transfer algorithms to replicate artistic techniques and aesthetics.
- 3D printing on canvas for physical reproduction of reconstructed artworks.
- Proprietary AI models trained on art historical data for accurate restoration.
- Digital reconstruction of pentimenti and underdrawings.
- Verification of authenticity through art historical research and expert consultation.