Self-initiated project

Memory Bridge

Restore the photograph. Keep the person you remember.

Project
Self-initiated · no client
My role
Product design & prototyping
Focus
Identity · intent · family review
Status
Working prototype
01 / Problem

A clearer face can still feel wrong.

Memory Bridge began with my own family photographs. A technically sharper image is not useful if the person looking at it no longer recognizes someone they know. I designed the flow around that judgment.

The person should remain recognizable, not just sharper.A portrait and its source frame the question of whether the person still feels familiar.
The person should remain recognizable, not just sharper.
02 / What exists today

The source is a photograph of someone people know.

The starting material is personal: a family photograph, supporting references and what someone remembers about the subject. A restoration proposes changes to that source. A new scene creates something that was never photographed.

The starting material includes what someone remembers.A family photograph sits among supporting references and the memory used to judge it.
The starting material includes what someone remembers.
03 / What’s wrong

One “generate” action hides two very different intentions.

Improving a source photograph and inventing a scene ask for different kinds of judgment. If the interface treats them as the same task, a convincing result can obscure what changed and what never happened.

Restoring a photograph and inventing a scene need different judgments.Two different image intentions are compressed into one action, making their boundary hard to read.
Restoring a photograph and inventing a scene need different judgments.
04 / What I explored

Separate the paths. Put the review before saving.

I worked through a reference setup, a restoration review and a separate creation flow. Person details and source photographs anchor the work; additional people, places and age variants belong to creating a scene.

The key choice was to let people inspect the instructions and compare the proposed restoration with the original before replacing the working photograph.

Separate the intentions and keep review before saving.Restoration and creation appear as distinct image paths, each with a deliberate review.
Separate the intentions and keep review before saving.
05 / Low-fi architecture

One reference set, two deliberate paths.

References inform both paths. Restoration leads to comparison and a save decision. Creation adds scene choices and produces a new interpretation.

THE ANCHOR

A person and their photographs

Source images and person details inform the work.

RESTORATION

Improve the source

Inspect the instructions.
Compare the proposed restoration.

Review → save
CREATION

Imagine a new scene

Choose people and place.
Review optional age variants.

An interpretation, not a record
06 / Wireframes

Put the family’s decision in the sequence.

Wireframe sequenceExplore the sequence
01 / 03

Give the person a clear reference

Person / ReferencesWireframe
Person
Primary photograph
Supporting references
Reference 2
Reference 3
Review restoration
01One primary reference anchors the work. Supporting photographs add context.

Reconstructed from the documented design; sample content is illustrative.

07 / Solution

Keep the original in view while judging the change.

Recognition cannot be delegated to a score.

People can inspect the instructions and compare the original with the proposed restoration. Save replaces the working photograph.

Compare the original photograph with the proposed restoration. These are actual project images; this page does not generate a new restoration.

Saving changes the working photo; the prototype does not provide a permanent source history.

Open the prototype ↗
08 / Research

The open question is recognition, not sharpness.

This is a self-initiated prototype developed from my own family photographs. The examples show what it produces and how review works; they are not evidence from a customer study.

I still need to test whether people recognize the subject, understand restoration versus invention, and know what Save changes. The current prototype replaces the working photograph and does not keep a permanent source history.

Family recognition and understanding of Save still need testing.Open questions surround the portrait: recognition, restoration versus invention, and what saving changes.
Family recognition and understanding of Save still need testing.
09 / Outcome

A working review flow. Family validation is still ahead.

Generation proposes a change; the person decides whether to save it.An original and a proposed restoration remain available for a human decision before saving.
Generation proposes a change; the person decides whether to save it.
Working prototype

Give people the final say before a photo changes.

My decision
I separated restoring a photograph from creating a new scene. Before saving a restoration, people can review the instructions and compare it with the original.
What changed
Generation proposes an image. Save replaces the working photograph. The person who knows the subject decides whether the result feels right.
Evidence
The working prototype includes this review step. Whether it improves recognition or trust still needs testing.
See how review works
1ONICA FM

Choose a perspective. Access is approved personally.

Public mode is open to everyone. Other perspectives open by invitation.

Level 010 XP

100 XP to the next level

Your progress stays in this browser.