AI Paper Atlas

Research map / Point clouds / Point cloud classification

BuildingGPT: Auto-Regressive Building Wireframe Reconstruction Model with Reinforcement Learning

Yuzhou Liu, Lingjie Zhu, Hanqiao Ye, Yujun Liu, Shangfeng Huang, Xiang Gao, Ruisheng Wang, Shuhan Shen

CVPR 2026 · Point clouds › Point cloud classification · approach: Point-language and open-vocabulary 3D

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
use hierarchical footprint-to-roof tokenization instead of vanilla z-y-x ordering (ablation)BuildingGPT pre-trained · MunichWF testWED (x10^-2): 1.39 -> 1.19Improved
use hierarchical tokenization (ablation)BuildingGPT pre-trained · MunichWF testEdge F1 (%): 91.9 -> 93.1Improved
add DPO post-training with preference pairs (ablation)BuildingGPT · MunichWF testWED (x10^-2): 1.19 -> 0.98Improved
add DPO post-training (ablation)BuildingGPT · MunichWF testEdge F1 (%): 93.1 -> 94.4Improved
replace diffusion/detection with auto-regressive wireframe generation (vs EdgeDiff)BuildingGPT · MunichWF testWED (x10^-2): 1.39 -> 0.98Improved
scale model from 129M to 730M parameters and data from 20% to 100% (ablation)BuildingGPT · MunichWF test cross-entropy losstest CE lossImproved
remove 25% of input points (ablation)BuildingGPT · MunichWF testWED (x10^-2): 0.98 -> 1.09No clear effect
remove 75% of input points (ablation)BuildingGPT · MunichWF testEdge F1 (%): 94.4 -> 76.6Hurt
add Gaussian noise 0.02 to input points (ablation)BuildingGPT · MunichWF testWED (x10^-2): 0.98 -> 1.33No clear effect
add Gaussian noise 0.05 to input points (ablation)BuildingGPT · MunichWF testEdge F1 (%): 94.4 -> 76.3Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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