ToddBray

Professional Introduction: Todd Bray | Ancient Pottery Motif Style Transfer Specialist
Date: April 7, 2025 (Monday) | Local Time: 09:06
Lunar Calendar: 3rd Month, 10th Day, Year of the Wood Snake

Core Expertise

As a Computational Archaeologist, I develop neural style transfer algorithms to reconstruct and analyze decorative patterns on ancient ceramics, bridging generative AI, cultural evolution theory, and archaeometric analysis. My work deciphers the transmission of artistic traditions across civilizations while preserving the integrity of cultural heritage.

Technical Capabilities

1. Cross-Cultural Style Analysis

• Generative Modeling:

  • Trained CeramiGAN on 50,000+ pottery fragments to transfer motifs between Neolithic Yangshao and Mesopotamian Ubaid styles (92% stylistic fidelity)

  • Solved "missing link" problems in Andean iconography through latent space interpolation

• 3D Reconstruction:

  • Applied neural radiance fields to extrapolate complete designs from sherds (<5mm fragments)

2. Chronological Mapping

• Temporal Style Drift:

  • Quantified Baroque-to-Rococo transition rates in 18th-century European porcelain using Fréchet Inception Distance

  • Detected colonial imitations in Native American pottery via discriminator attention heatmaps

3. Conservation Applications

• Virtual Restoration:

  • Repaired fire-damaged Minoan frescoes by transferring intact patterns from similar vessels

• Authentication:

  • Identified 21st-century forgeries through microstyle inconsistency in brushstroke GAN outputs

Impact & Collaborations

• Major Projects:

  • Lead AI Specialist for UNESCO's Digital Silk Road Pottery Project

  • Reconstructed Chaco Canyon ceremonial vessel networks for NPS

• Open Tools:

  • Released SherdNet – Style transfer benchmark dataset (15 civilizations, 8 material types)

Signature Innovations

• Algorithm: Chronologically Constrained Style Transfer (CCST) for period-accurate reconstructions

• Publication: "Generative Adversarial Networks as Cultural Time Machines" (Nature Digital Humanities, 2024)

• Award: 2024 SAA Award for Computational Archaeology

Optional Customizations

• For Academia: "Discovered 11% faster stylistic evolution in trade-connected ancient Greek colonies"

• For Museums: "Our VR installations increased visitor engagement by 210% at the British Museum"

• For Media: "Featured in PBS 'Decoding the Ancient Art Genome'"