Breathing modern life into an ancient art: AI and Rogan Kala

Rogan Kala and AI

Rogan Kala, a mesmerizing folk art from Gujarat's Kutch region, involves painting intricate designs on fabric using a special oil-based paint made from castor oil. The Khatri family applies the paint with a metal stylus, layers it on fabric, and reveals stunning patterns through a folding process. Passed down through generations, it's now on the brink of extinction.

But what if machine learning and AI could help preserve and evolve Rogan Kala without losing its soul?

Imagine using computer vision models like convolutional neural networks (CNNs) to analyze motifs, fractals, and symmetries. This database could generate new, authentic designs, suggesting variations that respect traditional rules while inspiring young artists.

Practically, simplified GANs or diffusion models could let artists simulate the rubbing technique on tablets, providing real-time feedback on paint flow and pattern accuracy. Workshops could use AR overlays to guide hands-on practice, bridging the gap for diaspora communities. Even e-commerce platforms could employ recommendation algorithms to match custom Rogan designs with buyers' tastes, boosting demand and funding artisans.

Using existing tech like Google's Arts & Culture experiments or AI pattern generators in tools like Adobe Sensei, makes this possible today. These systems, however, would require close collaboration with master artisans to encode cultural rules and ensure authenticity. By digitizing through AI, we could ensure Rogan Kala thrives for generations, blending heritage with innovation.

What do you think? Ready to code a prototype?