Loading page
About this tool
Upload a photo and simulate what your younger self would look like, along with retro polaroid frames and childhood traits summaries.
The Younger Version Generator takes a photo, finds the face with a 68-point landmark model running in your browser, and renders a nostalgic 'childhood self' polaroid by warming the exposure, softening the skin inside the detected face oval and adding rosy cheek tint. You pick one of three target stages — toddler (2-4), child (6-10) or teen (13-17) — and get back a framed image plus a playful profile with smooth-skin, cheek-roundness and eye-proportion figures. It is a photo filter and a novelty read, not a scientific reconstruction of how you actually looked.
Open Younger Version Generator on AltFTool — it loads instantly in your browser.
Pick a stage under "Select Younger Target Stage": "Toddler (2-4 yrs)", "Child (6-10 yrs)" or "Teen (13-17 yrs)".
Press "Select Photo" or drop an image on the dashed zone (max size 10MB); a TinyFaceDetector and 68-point landmark pass locates the face, then the canvas warms the exposure and blurs inside the face oval.
Compare "Your Photo" with the "Childhood Self" polaroid and its Smooth Skin Index, Cheek Roundness Density and Relative Eye Proportion bars, then press "Download" for younger-self-toddler.png or "Copy Report" for the plain-text version.
A TinyFaceDetector plus 68-point landmark pass finds the face, so the softening is masked to a face-shaped oval and the blush lands on the actual cheeks instead of the whole picture.
Toddler starts at 92% roundness and 98% skin smoothness, child at 82/92 and teen at 68/86, so each stage produces a visibly different read rather than one recycled result.
Download the result as a PNG named for the stage you chose, or copy a plain-text report listing the three indicators and the childhood personality and activity lines.
It detects your face with a lightweight in-browser landmark model, then edits the image on a canvas: exposure and warmth are lifted about 4%, an 8-pixel blur is blended into an oval over the detected face at up to 45% opacity, and a soft rose tint is drawn on each cheek at 12% alpha. The percentages shown alongside come from the age preset you selected, nudged slightly by your own landmark positions.
No. It is a stylised filter plus a randomised nostalgia profile, and it should be read as entertainment. Nothing in the tool has access to your actual childhood photos or any record of your appearance.
No. The image is read locally with FileReader and processed on a canvas in your browser; the face models load as static files and no picture leaves your device. Uploads are capped at 10 MB.
The tool falls back to a light global blur at 30% opacity and derives its seed from the file name and size instead of landmarks, so you still get a result — just without the face-shaped mask or the cheek blush. A clear, front-facing, well-lit photo gives the better output.