mmento-1
A camera for the sound of a moment
What if a camera could help us remember through sound?
- MY ROLE
- Device experience, interaction design and prototype development
- CONTEXT
- Independent exploration of musical memory
- STATUS
- Handheld prototype
A moment, held in your hand.
Press the shutter. The object captures an image as the starting point for a new piece of music.
From input to experience.
THE INTERACTION LOOP- 01
Capture
Press the shutter to hold an image of the moment.
- 02
Interpret
A local vision-language model proposes lyrics and style.
- 03
Compose
A music generation service creates the track.
- 04
Listen
The device plays it alongside the captured image.
Another sense of a moment
A photograph fixes what a moment looked like. I wanted to explore an audio snapshot: a piece of music prompted by the scene in front of you.
The experience starts with a familiar action. Press the physical shutter, capture an image, then let the object translate it into something you can listen to.
From a scene to a song
Once captured, the image freezes on screen. A local vision-language model interprets it and produces two structured outputs: lyrics and musical style.
These outputs go to a music generation service. When the track is ready, a callback brings it back to the device, which downloads and plays the music while overlaying the generated text on the image.
The whole loop, in your hand
The prototype brings together a Raspberry Pi-style setup, camera, screen, physical button, battery monitoring and local audio playback. I designed the device experience around the shutter rather than a sequence of software prompts.
Local image interpretation and remote music generation meet within the same object. The waiting and return of the track are part of the interaction the prototype explores.
A physical shutter connects an everyday gesture to an unfamiliar kind of memory.