Research map / 3D vision / 3D generation
MMOne: Representing Multiple Modalities in One Scene
Zhifeng Gu, Bing Wang
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 modality modeling module to joint-training baseline (ablation) | MM-J · RGB-Thermal-Language, RGB | PSNR: 22.32 -> 22.38 | No clear effect |
| add soft prune over hard prune (ablation) | MM + prune · RGB-Thermal-Language, RGB | PSNR: 22.67 -> 22.98 | Improved |
| add multimodal decomposition on top of soft prune (ablation) | MMOne · RGB-Thermal-Language, RGB | PSNR: 22.98 -> 23.19 | Improved |
| add multimodal decomposition on top of soft prune (ablation) | MMOne · RGB-Thermal-Language, Lang | mIoU: 47.0 -> 48.1 | Improved |
| add multimodal decomposition on top of soft prune (ablation) | MMOne · RGB-Thermal-Language, Thermal | PSNR: 23.99 -> 24.24 | Improved |
| add modality modeling module without disentangling (ablation) | MM-J · RGB-Thermal-Language, Lang | mIoU: 45.1 -> 45.3 | No clear effect |
| lower decomposition threshold from 0.0002 to 0.0001 (ablation) | MMOne · RGB-Thermal-Language | mIoU: 48.1 -> 47.1 | Hurt |
| raise decomposition threshold from 0.0002 to 0.0004 (ablation) | MMOne · RGB-Thermal-Language | mIoU: 48.1 -> 47.0 | Hurt |
| raise decomposition threshold from 0.0002 to 0.0003 (ablation) | MMOne · RGB-Thermal-Language | Gaussians (x10^4): 9.9 -> 10.1 | Hurt |
| replace joint training with MMOne | LangSplat joint training (LS-J) · RGB-Language | mIoU: 55.3 -> 56.6 | Improved |
| replace joint training with MMOne | LS-J · RGB-Language | RGB PSNR: 23.23 -> 24.35 | Improved |
| add thermal modality to RGB-Language | MMOne · Dimsum | mIoU: 57.1 -> 61.1 | Improved |
Closest papers by meaning
- MMIG-Bench: Towards Comprehensive and Explainable Evaluation of Multi-Modal Image Generation Models (NeurIPS 2025 · similarity 0.83)
- Universal Scene Graph Generation (CVPR 2025 · similarity 0.82)
- The Scene Language: Representing Scenes with Programs, Words, and Embeddings (CVPR 2025 · similarity 0.82)
- MMScan: A Multi-Modal 3D Scene Dataset with Hierarchical Grounded Language Annotations (NeurIPS 2024 · similarity 0.81)
- MMGDreamer: Mixed-Modality Graph for Geometry-Controllable 3D Indoor Scene Generation (AAAI 2025 · similarity 0.81)
- SceneDecorator: Towards Scene-Oriented Story Generation with Scene Planning and Scene Consistency (NeurIPS 2025 · similarity 0.80)
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.
