Research map / Label-efficient and robust learning / Domain adaptation and generalization
Walking Your LiDOG: A Journey Through Multiple Domains for LiDAR Semantic Segmentation
Cristiano Saltori, Aljosa Osep, Elisa Ricci, Laura Leal-Taixé
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 |
|---|---|---|---|
| add the BEV auxiliary segmentation task | MinkowskiNet segmenter · Synth4D-KITTI -> SemanticKITTI | mIoU: 24.69 -> 44.18 | Improved |
| add the BEV auxiliary task | MinkowskiNet segmenter · Synth4D-KITTI -> nuScenes | mIoU: 20.62 -> 37.14 | Improved |
| add the BEV auxiliary task | MinkowskiNet segmenter · SemanticKITTI -> nuScenes | mIoU: 26.53 -> 34.88 | Improved |
| add the BEV auxiliary task | MinkowskiNet segmenter · nuScenes -> SemanticKITTI | mIoU: 29.55 -> 41.22 | Improved |
| add the BEV auxiliary task | MinkowskiNet segmenter · Synth4D-nuScenes -> SemanticKITTI | mIoU: 19.71 -> 34.79 | Improved |
| add the BEV auxiliary task | MinkowskiNet segmenter · Synth4D-nuScenes -> nuScenes | mIoU: 24.57 -> 33.78 | Improved |
| add the BEV auxiliary task (multi-source training) | MinkowskiNet segmenter · multi-source -> SemanticKITTI | mIoU: 31.82 -> 42.44 | Improved |
| add the BEV auxiliary task (multi-source training) | MinkowskiNet segmenter · multi-source -> nuScenes | mIoU: 25.60 -> 40.23 | Improved |
| train on two synthetic source domains instead of one | Source model · Synth4D -> SemanticKITTI | mIoU: 24.69 -> 31.82 | Improved |
| train on two synthetic source domains instead of one | Source model · Synth4D -> nuScenes | mIoU: 20.62 -> 25.60 | Improved |
| replace the BEV decoder with an extra 3D head and ensemble predictions (Double) (ablation) | LiDOG · Synth4D-KITTI -> SemanticKITTI and nuScenes | mIoU | Hurt |
| increase BEV area from 10x10 to 60x60 m (ablation) | LiDOG · Synth4D-KITTI -> Real | mIoU | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- CoSMix — CoSMix: Compositional Semantic Mix for Domain Adaptation in 3D LiDAR Segmentation (ECCV 2022) · better on 2 of 2 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
- Single Domain Generalization for LiDAR Semantic Segmentation (CVPR 2023 · similarity 0.95)
- Toward Enhanced Representation Learning for Single-Source Domain Generalization in LiDAR Semantic Segmentation (TPAMI 2026 · similarity 0.94)
- Domain Generalization of 3D Semantic Segmentation in Autonomous Driving (ICCV 2023 · similarity 0.91)
- LESS: Label-Efficient Semantic Segmentation for LiDAR Point Clouds (ECCV 2022 · similarity 0.91)
- MemorySeg: Online LiDAR Semantic Segmentation with a Latent Memory (ICCV 2023 · similarity 0.91)
- PC-BEV: An Efficient Polar-Cartesian BEV Fusion Framework for LiDAR Semantic Segmentation (AAAI 2025 · similarity 0.89)
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.
