Human-in-the-loop AI framework reconstructs color in historical lenticular films

Researchers have developed a new human-in-the-loop deep learning framework to improve the color reconstruction of historical lenticular films, such as those made with the Kodacolor process. This framework allows experts to interactively refine lenticule boundaries, embedding their knowledge into the model to enhance robustness against challenges like curved lenticules or low contrast. The system merges reconstructed color information with the original film scan's luminance to preserve image details and texture, successfully producing high-quality, exhibitable color reconstructions where previous automated methods failed. AI

IMPACT Improves AI's ability to handle specialized historical data reconstruction tasks.

RANK_REASON Academic paper detailing a new deep learning framework. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• deep-doLCE
• doLCE
• Kodacolor
• Saptarshi Neil Sinha

• paper
• product
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →