Google DeepMind Uses AI to Recreate Couple's Unrecorded Memory in 'Love, Rendered'
Google DeepMind detailed how it used image restoration and performance capture AI to help a couple recreate a cherished memory that was never photographed.
Quick answer
What did Google DeepMind announce about the documentary 'Love, Rendered'?
Google DeepMind says it collaborated with filmmakers on 'Love, Rendered,' a documentary that used generative image restoration and performance capture models to recreate a memory for Burt and Ethelle Shatz, a couple married over 70 years, that was never photographed or filmed. Google also notes users can restore and colorize family photos with the Gemini app.
Key takeaways
- Google DeepMind collaborated with filmmakers Liz Garbus, Dan Cogan, and Darren Aronofsky's Primordial Soup on a documentary short called 'Love, Rendered.'
- The film follows Burt and Ethelle Shatz, married more than 70 years, and uses AI to recreate the day they met at a student co-op in Cleveland, an event that was never photographed or filmed.
- Google DeepMind says its team used a two-part technical approach: generative image restoration to repair old black-and-white photos, and performance capture models to map the couple's present-day mannerisms onto their younger likenesses.
- Google DeepMind engineer Michael Chang says Ethelle acted as an active co-creator during the process, correcting details like the curve of a staircase or the shape of a shoe heel.
- Google says people can use the Gemini app to restore and colorize their own family photographs.
Google DeepMind said it worked with filmmakers to produce "Love, Rendered," a documentary short that uses artificial intelligence to help a couple recreate a cherished memory that was never captured on film or in photographs.
The project, described in a post by Google DeepMind engineer Michael Chang, centers on Burt and Ethelle Shatz, a couple married for more than 70 years. According to Google, the film follows the couple as they navigate Burt's cognitive decline, including the fading of one of their most treasured memories: the day they met at a student co-op in Cleveland. Because that day was never photographed or filmed, Google says it existed only in Burt and Ethelle's minds.
"Love, Rendered" was directed by Liz Garbus and produced alongside Dan Cogan and Darren Aronofsky, Google said. The film was created through a collaboration between Google DeepMind and Primordial Soup, described as Aronofsky's creative venture.
Two-part technical approach
Google DeepMind said bridging the gaps in the couple's missing visual record required a two-part technical process intended to preserve what the company calls emotional truth:
- Image restoration: According to Google, the team used generative models to restore black-and-white photographs of Burt and Ethelle from their youth, which it says helped ensure later recreations stayed faithful to their subjects.
- Pose and performance control: Google DeepMind said engineers used performance capture models to map the couple's present-day micro-mannerisms — including the tilt of Burt's head, a hesitation in his speech, and the crinkle around his eyes — onto their younger likenesses.
Google said combining these two methods let the team "intertwine Burt and Ethelle's past and their present," producing a recreated memory the couple described as feeling authentic.
Chang wrote that Ethelle worked directly alongside the technical team as what Google DeepMind calls an active co-creator, correcting specific visual details such as the curve of a staircase or the shape of a shoe heel to guide the AI output.
Chang also wrote that the project had personal resonance for him, describing testing the same image restoration and video animation tools on his own family's old photographs after his grandfather experienced memory loss following a stroke.
Google said the underlying restoration and colorization capability used in the project is available to the public through the Gemini app, where users can apply it to their own family photographs.
Google also pointed to a companion explanation of the technical workflow from colleague Jess Gallegos, referenced in Chang's post.
Source: Google, "Recreating a 70-year love story frame by frame," published September 9, 2026.
Frequently asked questions
- What is 'Love, Rendered'?
- It is a documentary short, directed by Liz Garbus and produced with Dan Cogan and Darren Aronofsky, about Burt and Ethelle Shatz, a couple married over 70 years, made in collaboration with Google DeepMind and Aronofsky's Primordial Soup.
- What memory did the film recreate?
- Google DeepMind says the film recreates the day Burt and Ethelle met at a student co-op in Cleveland, a memory that existed only in their minds because it was never photographed or filmed.
- What AI techniques did Google DeepMind use?
- According to Google DeepMind, the team combined generative image restoration on old black-and-white photos with performance capture models that mapped the couple's current micro-mannerisms onto their younger likenesses.
- Can the public use this technology?
- Google says people can use the Gemini app to restore and colorize their own family photographs.