The disappearance of the Mona Lisa - and its unusual retrieval

In the digital age, deleted data does not always mean lost data. Carve-DL, a cutting-edge research project at the German Research Center for Artificial Intelligence (DFKI), explores how deep learning can enhance the reconstruction of fragmented and erased digital information.
This video illustrates the methodology behind Carve-DL through a fictional case: the theft of the Mona Lisa and the challenge of recovering digital evidence that has been deliberately deleted. By leveraging AI-powered classification, clustering, and reassembly techniques, Carve-DL enables forensic specialists to reconstruct lost data with unprecedented accuracy.

🔬 Developed by the Generative & Transparent AI (GeT-AI) research team, led by Dr. Tobias Wirth, Senior Consultant in the Department of Smart Data & Knowledge Services, this project represents a major step forward in AI-assisted digital forensics.

📌 Learn more about Carve-DL: https://www.carve-dl.de
🔍 Discover DFKI’s research: https://www.dfki.de