Open source · GPL-3.0

Stereoscopic visualization

pAR3D

A Windows experiment that turns desktop or camera feeds into side-by-side 3D imagery using AI depth estimation.

Illustration of left and right stereoscopic views; not a pAR3D screenshot.
Illustration of left and right stereoscopic views; not a pAR3D screenshot.

What it is useful for.

  • Desktop or camera capture and side-by-side output
  • Adjustable depth, scale and performance modes
  • Optional CUDA acceleration with CPU operation supported
PlatformWindows / Python
AvailabilityOpen source · GPL-3.0

Current scope & limitations

Designed for experimentation and creative visualization. Performance depends on the capture workload and hardware; estimated depth is not a measured 3D reconstruction.

Keep exploring.

Compare the rest of the workshop, or contact Edwin about integration, support, and commercial work.