Research map / Point clouds / Segmentation and detection
GIPSO: Geometrically Informed Propagation for Online Adaptation in 3D LiDAR Segmentation
Cristiano Saltori, Evgeny Krivosheev, Stéphane Lathuilière, Nicu Sebe, Fabio Galasso, Giuseppe Fiameni, Elisa Ricci, Fabio Poiesi
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 changed | Where | Measured | Effect |
|---|---|---|---|
| increase propagation size K from 1 to 10 (ablation) | GIPSO · Synth4D to SemanticKITTI | mIoU improvement over Source: +4.27 -> +4.31 | Improved |
| increase propagation size K from 10 to 50 (ablation) | GIPSO · Synth4D to SemanticKITTI | mIoU improvement over Source: +4.31 -> +3.99 | Hurt |
| increase propagation size K from 10 to 100 (ablation) | GIPSO · Synth4D to SemanticKITTI | mIoU improvement over Source: +4.31 -> +3.91 | Hurt |
| increase propagation size K from 1 to 5 (ablation) | GIPSO · Synth4D to SemanticKITTI | mIoU improvement over Source: +4.27 -> +4.26 | No clear effect |
| increase time window w from 1 to 6 (ablation) | GIPSO temporal consistency loss · Synth4D to SemanticKITTI | mIoU improvement over Source: +3.35 -> +4.40 | Improved |
| increase time window w from 6 to 8 (ablation) | GIPSO temporal consistency loss · Synth4D to SemanticKITTI | mIoU improvement over Source: +4.40 -> +4.21 | Hurt |
| add GIPSO adaptive selection, dropout pseudo-labeling and no source momentum update (ablation) | ProDA* · Synth4D to SemanticKITTI | mIoU improvement over Source: -32.63 -> +1.48 | Improved |
| add GIPSO uncertainty-based adaptive pseudo-label selection (ablation) | CBST* · Synth4D to SemanticKITTI | mIoU improvement over Source: +0.28 -> +1.07 | Improved |
| add GIPSO adaptive selection and geometric propagation instead of spatial nearest neighbor (ablation) | TPLD* · Synth4D to SemanticKITTI | mIoU improvement over Source: 0.56 -> 1.38 | Improved |
Closest papers by meaning
- CoSMix: Compositional Semantic Mix for Domain Adaptation in 3D LiDAR Segmentation (ECCV 2022 · similarity 0.87)
- D^3CTTA: Domain-Dependent Decorrelation for Continual Test-Time Adaption of 3D LiDAR Segmentation (CVPR 2025 · similarity 0.86)
- Test-Time Training for LiDAR Semantic Segmentation under Corruption via Geometric Inlier Discrimination (CVPR 2026 · similarity 0.85)
- LiDAR-guided Geometric Pretraining for Vision-Centric 3D Object Detection (IJCV 2025 · similarity 0.84)
- An Empirical Study of the Generalization Ability of Lidar 3D Object Detectors to Unseen Domains (CVPR 2024 · similarity 0.82)
- GBlobs: Explicit Local Structure via Gaussian Blobs for Improved Cross-Domain LiDAR-based 3D Object Detection (CVPR 2025 · similarity 0.82)
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.
