// projects

PatternBuffer

Active

Turns a photo or a line of text into living 3D particles, and hands back a loop that closes on itself. The depth model runs on your device, so the photo never leaves it.

*The editor: a photo as particles, with Depth and Looks to hand*

*Relief turns the picture so the depth reads, here in dark mode*

*Export a video, GIF, still, sticker or link, at the stage's exact frame*

1 / 3

// summary

Most photo-to-3D tools hand you a gimmick you see once. PatternBuffer is built for the feed: drop in a photo or type a line, and it becomes a cloud of particles that sways, bursts, assembles out of dust or breathes, on a loop that closes exactly on itself. What you see on the stage is the file you get, in 9:16, square, 4:5 or Spotify Canvas.

Depth comes from the photo’s brightness, or from an AI model that runs on your own device after a one-time download you agree to first. A second model finds the subject, so it can lift off the background. Nothing about the photo leaves the phone unless you share it.

Export a video, a GIF, a still or a sticker, or share a link anyone can play with. There are no accounts: your browser remembers what you shared and can take it down. The name is a Star Trek joke. The pattern buffer is the bit of the transporter that holds you mid-beam.

// tech stack

Frontend

Three.js WebGL React 19 Tailwind CSS 4 Base UI

Language

TypeScript

AI / ML

Transformers.js ONNX Runtime Web Hugging Face

Infrastructure

Cloudflare Workers Cloudflare R2

Testing

Playwright

Tooling

Vite