Description
ROOM is a spatial memory app that rebuilds meaningful places from the photos and videos already scattered across your camera roll. Its author made it as a gift to their future self. Instead of treating photos as isolated moments, ROOM uses multimodal AI and spatial reconstruction to recognize the same place across years, rebuild it as a navigable 3D space, and return each memory to where it happened.
A room is more than a container for memories. Its walls, furniture, light and objects, even the spots that later sit empty, become part of what we remember. The prototype preserves the author's own room: a place they will eventually leave, and one that still carries traces of their dog, who is no longer here. Selecting an old photo brings the view back to where it was taken, at the same camera angle, and a THEN / NOW interaction lines the past up with the present.
ROOM keeps a record not only of what happened, but of where life happened. Today the prototype uses real spatial scans and aligned photos; the longer-term vision is for AI to keep discovering places in a personal photo library and build a living spatial memory over the years.
With HeyGears 3D printing, the author also turns a simplified room model into a physical keepsake. It preserves the room's spatial structure, while the richer photos, time layers and memories stay in the linked digital world: an object that can sit on a desk and reopen the room years later.
How Tripo was used
Tripo turned personal objects from the author's own photos into 3D assets, exploring how things that are no longer in the room can be restored in space rather than merely shown as flat photos. These assets were tested inside ROOM's spatial-memory workflow, connecting photo understanding, 3D reconstruction and the interactive room.
