AI Paper Atlas

Research map / 3D vision / 3D generation

PrITTI: Primitive-based Generation of Controllable and Editable 3D Semantic Urban Scenes

Christina Ourania Tze, Daniel Dauner, Yiyi Liao, Dzmitry Tsishkou, Andreas Geiger

CVPR 2026 · 3D vision › 3D generation · approach: Text-to-3D scene generation · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add latent split in the LVAE (ablation)PrITTI LVAE · KITTI-360 LVAE ablationRaster MSE (x10^-2): 3.55 -> 0.75Improved
add latent split in the LVAE (ablation)PrITTI LVAE · KITTI-360 LVAE ablationAP3D: 53.78 -> 62.12Improved
add latent split in the LVAE (ablation)PrITTI LVAE · KITTI-360 LVAE ablationAP3D@50: 39.09 -> 46.96Improved
joint training with ground raster branch (ablation)PrITTI LVAE · KITTI-360 LVAE ablationAP3D: 60.28 -> 62.12Improved
joint training with object branch (ablation)PrITTI LVAE · KITTI-360 LVAE ablationRaster MSE (x10^-2): 0.59 -> 0.75Hurt
increase SemCity decoder query points from 400K to 1MSemCity · KITTI-360 voxel reconstructionmIoU: 70.52 -> 93.81Improved
reduce voxel resolution then upsampleSemCity · KITTI-360 voxel reconstructionmIoU: 70.52 -> 50.52Hurt
primitive representation instead of voxelsPrITTI · KITTI-360 voxel reconstructionSize (MB): 8.00 -> 2.52Improved
scale DiT-B to DiT-LPrITTI · KITTI-360 3D scene generationFID: 73.952 -> 71.028Improved
scale DiT-L to DiT-XLPrITTI · KITTI-360 3D scene generationFID: 71.028 -> 71.735No clear effect
PrITTI instead of XCube Level 2PrITTI DiT-B · KITTI-360 3D scene generationFID: 94.822 -> 73.952Improved
Cholesky encoding instead of quaternions (ablation)PrITTI object encoding · synthetic vehicle primitivesIoU3DImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy