AI Paper Atlas

Research map / Point clouds / Segmentation and detection

High-Quality Pseudo-Labeling for Point Cloud Segmentation With Scene-Level Annotation

Lunhao Duan, Shanshan Zhao, Xingxing Weng, Jing Zhang, Gui-Song Xia

TPAMI 2025 · Point clouds › Segmentation and detection · approach: Point cloud semantic segmentation · 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 cross-modal feature guidance (3D contrastive loss) (ablation)baseline · ScanNet train pseudo-labels 3DmIoU: 25.1 -> 36.8Improved
add region-point consistency without CMG (ablation)baseline · ScanNet train pseudo-labels 3DmIoU: 25.1 -> 29.6Improved
add region-point consistency (3D) on top of CMG (ablation)CMG · ScanNet train pseudo-labels 3DmIoU: 36.8 -> 47.7Improved
add 2D branch region-point consistency loss (ablation)CMG + RPC 3D · ScanNet train pseudo-labels 3DmIoU: 47.7 -> 48.2Improved
add Mix3D augmentation (ablation)full method · ScanNet train pseudo-labels 3DmIoU: 48.2 -> 50.1Improved
add Mix3D augmentation (ablation)full method · S3DIS train pseudo-labels 3DmIoU: 50.5 -> 51.6Improved
use contrastive distillation instead of MSE for 2D-to-3D (ablation)CMG · ScanNet train pseudo-labels 3DmIoU: 32.6 -> 36.8Improved
use contrastive distillation instead of CAM guidance (ablation)CMG · ScanNet train pseudo-labels 3DmIoU: 32.7 -> 36.8Improved
use unsupervised regions instead of point-wise consistency (ablation)RPC · ScanNet train pseudo-labelsmIoU: 44.5 -> 48.2Improved
use adaptive per-category threshold instead of fixed (eta 0.8) (ablation)RPC · ScanNet train pseudo-labelsmIoU: 46.6 -> 48.2Improved
use 6 image views instead of none, versus 12 views (ablation)CMG · ScanNet train pseudo-labels 3DmIoU: 25.1 -> 36.8; 36.8 -> 33.1Mixed
initialize 2D backbone with ImageNet weights (ablation)ResNet50 · ScanNet train pseudo-labels 3DmIoU: 40.0 -> 48.2Improved

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