Research map / 3D vision / 3D generation
OccScene: Semantic Occupancy-based Cross-task Mutual Learning for 3D Scene Generation
Bohan Li, Xin Jin, Jianan Wang, Yukai Shi, Yasheng Sun, Xiaofeng Wang, Zhuang Ma, Baao Xie, Chao Ma, Xiaokang Yang, Wenjun Zeng
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 joint diffusion scheme (attached perception model) (ablation) | OccScene · SemanticKITTI ablation | FID: 28.52 -> 19.86 | Improved |
| add joint diffusion scheme (attached perception model) (ablation) | OccScene · SemanticKITTI ablation | mIoU: 12.94 -> 14.98 | Improved |
| add Mamba-based dual alignment (ablation) | OccScene · SemanticKITTI ablation | FVD: 162.04 -> 113.28 | Improved |
| add Mamba-based dual alignment (ablation) | OccScene · SemanticKITTI ablation | mIoU: 13.41 -> 14.98 | Improved |
| use Mamba instead of GRU encoding in alignment (ablation) | OccScene · SemanticKITTI ablation | FID: 24.54 -> 19.86 | Improved |
| use Mamba instead of attention encoding in alignment (ablation) | OccScene · SemanticKITTI ablation | Time: 4.09 -> 2.76 | Improved |
| apply MDA along depth dimension of occupancy (ablation) | OccScene · SemanticKITTI ablation | FID: 22.12 -> 19.86 | Improved |
| apply MDA along temporal dimension of video latent (ablation) | OccScene · SemanticKITTI ablation | FID: 23.02 -> 19.86 | Improved |
| attach gradients between generative and perception models (ablation) | MonoScene on NYUv2 · NYUv2 | mIoU: 27.47 -> 29.78 | Improved |
| attach gradients between generative and perception models (ablation) | NDC-Scene on NYUv2 · NYUv2 | mIoU: 30.20 -> 32.12 | Improved |
| attach gradients between generative and perception models (ablation) | ISO on NYUv2 · NYUv2 | mIoU: 32.09 -> 33.92 | Improved |
| train MonoScene with OccScene synthetic data instead of MagicDrive data | MonoScene · NuScenes-Occupancy | mIoU: 7.2 -> 10.2 | Improved |
Later papers that compared with this method
1 later accepted paper reports OccScene as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- OmniNWM: Unifying the State-Action-Reward Triad for Closed-Loop Panoramic Driving Navigation World Models (ECCV 2026 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- MonoScene — MonoScene: Monocular 3D Semantic Scene Completion (CVPR 2022) · better on 3 of 3 shared results
- SemCity — SemCity: Semantic Scene Generation with Triplane Diffusion (CVPR 2024) · better on 1 of 1 shared result
- MagicDrive — MagicDrive: Street View Generation with Diverse 3D Geometry Control (ICLR 2024) · better on 1 of 1 shared result
- OccFormer — OccFormer: Dual-path Transformer for Vision-based 3D Semantic Occupancy Prediction (ICCV 2023) · better on 1 of 1 shared result
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
- RadOcc: Learning Cross-Modality Occupancy Knowledge through Rendering Assisted Distillation (AAAI 2024 · similarity 0.89)
- OccMamba: Semantic Occupancy Prediction with State Space Models (CVPR 2025 · similarity 0.88)
- LOMA: Language-assisted Semantic Occupancy Network via Triplane Mamba (AAAI 2025 · similarity 0.88)
- OctOcc: High-Resolution 3D Occupancy Prediction with Octree (AAAI 2024 · similarity 0.87)
- Rethinking Temporal Fusion with a Unified Gradient Descent View for 3D Semantic Occupancy Prediction (CVPR 2025 · similarity 0.87)
- LowRankOcc: Tensor Decomposition and Low-Rank Recovery for Vision-based 3D Semantic Occupancy Prediction (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.
