Tree-D Studio

Canvas to Relief

AI‑assisted preservation

Transform paintings into tactile 3D reliefs

Tree‑D converts historic canvases into depth‑aware 3D objects. We analyze brushwork with Marigold normal estimation, translate luminance into displacement, and craft roughness maps for a realistic museum‑grade surface.

Featured Relief

01

Curatorial Search

Browse the Met collection and select works for relief conversion.

02

Studio View

Inspect each painting in a gallery‑grade 3D viewer with raking light.

03

Experimental Demo

Test custom IDs or upload your own image for rapid prototyping.

Technology

Depth inference meets conservation craft

We combine AI‑generated normals with procedural smoothing to preserve pigment flow, then translate it into displacement for genuine relief. Roughness maps retain varnish and glaze variation.

Normal EstimationMarigold
Surface ReliefDisplacement
Material FinishRoughness

AR Placement

View the relief in your room (iPhone)

  1. Find a painting from the archive and export the USDZ file.
  2. Download the USDZ to your iPhone (Files app).
  3. Tap the file to open Quick Look, then place the painting using AR.

The AR result should resemble the sample below (Thank you Verdio Wong =D).

Painting placed in room using AR