In collaboration with filmmaker Liz Garbus and Primordial Soup, Google DeepMind utilized generative AI models to reconstruct an unphotographed meeting between Burt and Ethelle Shatz for the documentary short "Love, Rendered." The project addresses cognitive decline by recreating a specific memory that existed only in the couple's minds.
The technical approach involved two main components: using generative models to restore black-and-white photos of the couple from their youth, and employing performance capture models to map Burt’s present-day micro-mannerisms onto their younger likenesses. This combination allowed Ethelle to act as a co-creator, correcting details like staircase curves and shoe shapes to ensure emotional truth.
The filmmakers leveraged these tools to explore reminiscence therapy, aiming to rekindle memories through sensory cues. The resulting animation was described by the couple as authentic, demonstrating how AI can help preserve fading personal history.