ComfyUI-DreamScene360 builds 3D room layouts from one photo

A floating geometric point cloud that reconstructs the room into delicate spatial lines and dots.

ComfyUI-DreamScene360 is a new custom node that converts a single 360-degree panoramic image into a three-dimensional point cloud. The tool measures distances across the scene to build spatial layouts without requiring multiple photographs.

Creator jfirma1 packaged recent academic research into a simple workflow format. This removes the need for outside accounts or extra generation steps.

Turning flat panoramas into spatial data

  • Processes standard wide-angle images into 3D point cloud format.
  • Aligns overlapping depth maps to reduce visible seam artifacts.
  • Outputs standard .ply file formats for direct use in other 3D tools.
  • Caches processing results to skip repeated calculations on the same scene.

Designers visualizing rooms or testing digital sets can use this workflow to quickly check camera angles. Feeding one wide-angle photo into the system produces a structured dataset ready for editing and lighting changes.

Setup notes and performance realities

The underlying system relies on specialized programming files that must be built locally on your machine. To avoid software conflicts, the included setup tool locks the project to an older version, which stops hidden crashes that newer updates often cause.

Users should expect processing to take five to fifteen minutes depending on hardware strength and image settings.

'Think of this as a starting point for blocking and staging, not a finished environment,'

noted the creator in a community post. Moving too far from the original camera position creates visible gaps, so keeping expectations realistic ensures smoother results.

Download the ComfyUI-DreamScene360 package from GitHub.