Research map / 3D vision / 3D generation
MergeOcc: Bridge the Domain Gap between Different LiDARs for Robust Occupancy Prediction
Zikun Xu, Shaobing Xu
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 |
|---|---|---|---|
| align point cloud range to minimum range (ablation) | MergeOcc · OpenOccupancy-nuScenes | geometric IoU: 25.4 -> 38.2 | Improved |
| align point cloud range to minimum range (ablation) | MergeOcc · SemanticKITTI | geometric IoU: 36.7 -> 70.0 | Improved |
| align range to maximum range instead of none (ablation) | MergeOcc · OpenOccupancy-nuScenes | geometric IoU: 25.4 -> 16.5 | Hurt |
| add geometric statistic alignment alone (dataset-specific norm) (ablation) | MergeOcc · OpenOccupancy-nuScenes | geometric IoU: 25.4 -> 14.7 | Hurt |
| add geometric statistic alignment on top of range alignment (ablation) | MergeOcc · OpenOccupancy-nuScenes | geometric IoU: 38.2 -> 38.4 | No clear effect |
| add semantic label mapping (ablation) | MergeOcc · OpenOccupancy-nuScenes | semantic mIoU: 21.4 -> 22.9 | Improved |
| add semantic label mapping (ablation) | MergeOcc · SemanticKITTI | geometric IoU: 69.9 -> 70.1 | No clear effect |
| use geometric alignment only in the backbone instead of all norm layers (ablation) | MergeOcc · OpenOccupancy-nuScenes | geometric IoU: 37.0 -> 38.4 | Improved |
| range alignment under single-dataset training (ablation) | MergeOcc · OpenOccupancy-nuScenes | geometric IoU: 30.9 -> 27.4 | Hurt |
| pre-train on SK then fine-tune on OO-nu (ablation) | L-CONet · OpenOccupancy-nuScenes | geometric IoU: 30.1 -> 26.3 | Hurt |
| pre-train on OO-nu then fine-tune on SK (ablation) | L-CONet · SemanticKITTI | geometric IoU: 55.6 -> 53.3 | Hurt |
| train on both datasets with MergeOcc instead of single dataset | L-CONet · SemanticKITTI | geometric IoU: 55.6 -> 70.1 | Improved |
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.
- SurroundOcc — SurroundOcc: Multi-camera 3D Occupancy Prediction for Autonomous Driving (ICCV 2023) · 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
- SDGOCC: Semantic and Depth-Guided Bird's-Eye View Transformation for 3D Multimodal Occupancy Prediction (CVPR 2025 · similarity 0.89)
- DIO: Decomposable Implicit 4D Occupancy-Flow World Model (CVPR 2025 · similarity 0.89)
- Towards foundational LiDAR world models with efficient latent flow matching (NeurIPS 2025 · similarity 0.88)
- EvOcc: Accurate Semantic Occupancy for Automated Driving Using Evidence Theory (CVPR 2025 · similarity 0.87)
- UnO: Unsupervised Occupancy Fields for Perception and Forecasting (CVPR 2024 · similarity 0.87)
- Accurate Training Data for Occupancy Map Prediction in Automated Driving Using Evidence Theory (CVPR 2024 · similarity 0.87)
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.
