KIRI Engine 4.0: Major Updates in Photogrammetry and AI-Powered PBR

September 8, 2025

Photo Scan mode in KIRI Engine is powered by KIRI’s cutting-edge photogrammetry technologies, it has been one of the app’s most popular features, but the team knew it could be better. In this latest update, they are focusing on improving the photogrammetry and the Photo Scan Mode in KIRI Engine.

Limitations of previous Photo Scan Mode

Photogrammetry is a proven and widely adopted technique, but its performance depends heavily on the surface texture of the object.

Subjects with clear patterns and rich details typically generate stable, accurate scans, whereas smooth or low-texture objects may lack sufficient feature points, often resulting in incomplete or unstable reconstructions.

Photo Scan’s Major improvements in KIRI Engine 4.0

Lower Texture Requirements

Even low-texture objects can now be captured more reliably, reducing common failures seen in previous versions.

More Stable Results

The new Photo Scan Mode delivers improved surface quality, producing cleaner and more accurate 3D reconstructions.

5x Faster Processing

The processing speed in Photo Scan has increased by up to five times, it has been reduced from nearly 10 minutes to around 2 minutes, significantly improving efficiency.

Expanded Photo Upload Limits

Providing more high-quality images generally leads to better reconstruction. With KIRI Engine 4.0, the upload limits have been increased:

  • KIRI Engine Basic: 100 → 150 photos
  • KIRI Engine Pro: 300 → 500 photos

This enables capturing more perspectives and achieving higher-quality scans.

New AR Mode

When enabled, AR Mode shows each captured photo as a 3D camera icon in space, helping users track coverage and ensure all angles are captured.

Unlimited Scans and Exports for Free Users

Previously, KIRI Engine Basic users could enjoy unlimited scans with Photo Scan but were limited to three exports per week. With this update, free users can now also enjoy unlimited exports, making KIRI Engine a truly sustainable tool for long-term use.

AI-powered PBR Material Capturing

KIRI Engine 4.0 introduces a new PBR training algorithm. For the first time in 3D scanning, the KIRI Engine can generate professional-grade PBR materials using a diffusion model — the same technology that powers modern AI image generation. This breakthrough enables capturing more realistic and physically accurate textures directly from scans.

Full-res In-app Rendering & Texture toggle

Previously, KIRI Engine displayed only low-poly previews on mobile devices. With KIRI Engine 4.0, a new compression algorithm enables full-resolution 3D models to be rendered directly on mobile, and users can now freely toggle textures on or off for better mesh inspection.

Learn More & Join the Community

To find out more, visit us at https://www.kiriengine.app/

Join other KIRI Engine users on our Discord

Follow KIRI Engine on Instagram

Follow KIRI Engine on Twitter/X

Try the KIRI Engine now for free:

Download KIRI Engine on Google Play Store: https://play.google.com/store/apps/details?id=com.kiriengine.app

Download KIRI Engine on App Store: https://apps.apple.com/us/app/kiri-engine-3d-scanner-lidar/id1577127142

KIRI Engine WEB Version: https://www.kiriengine.app/webapp/

Brought to you by KIRI Innovations:
This article is part of the befores & afters VFX Insight series. If you’d like to promote your VFX/animation/CG tech or service, you can find out more about the VFX Insight series here.

Leave a Reply

Don't Miss

KIRI Engine’s New iPhone Update Uses AI To Fix LiDAR’s Quality Limitations

KIRI is excited to announce KIRI Engine Version 3.14.

The Evolution of 3D Gaussian Splatting in Blender: A Look at the Latest 3DGS Render Addon Update

This article explores the latest update of the 3DGS Render

Why 3D Gaussian Splatting Just Got Better with KIRI Engine Version 3.13

New 3D Gaussian Splatting tech, a website redesign, and more.

Discover more from befores & afters

Subscribe now to keep reading and get access to the full archive.

Continue reading