Research map / 3D vision / 3D generation
YouTube-Occ: Learning Indoor 3D Semantic Occupancy Prediction from YouTube Videos
Haoming Chen, Lichen Yuan, TianFang Sun, Jingyu Gong, Zhizhong Zhang, Xin Tan, Yanyun Qu, Yuan Xie
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 YouTube-Occ pre-training | Symphonies · NYUv2 | mIoU: 29.70 -> 31.65 | Improved |
| add YouTube-Occ pre-training | Symphonies · Occ-ScanNet | mIoU: 47.71 -> 49.38 | Improved |
| add YouTube-Occ pre-training | EmbodiedOcc · NYUv2 | mIoU: 19.14 -> 20.18 | Improved |
| add YouTube-Occ pre-training | EmbodiedOcc · Occ-ScanNet | mIoU: 44.69 -> 46.17 | Improved |
| add YouTube-Occ pre-training | MonoScene · NYUv2 | mIoU: 26.09 -> 27.02 | Improved |
| train Symphonies directly on pseudo-labels (zero-shot) | Symphonies · Occ-ScanNet | mIoU: 8.41 vs 10.06 raw pseudo-labels | Hurt |
| joint-train with YouTube-Occ instead of NYUv2 | Symphonies · Occ-ScanNet | mIoU: 40.17 -> 37.38 | Hurt |
| pre-train on YouTube-Occ with 5% labels | Symphonies · NYUv2 few-shot | mIoU: 15.52 -> 18.81 (+3.29) | Improved |
| pre-train on YouTube-Occ with 10% labels | Symphonies · NYUv2 few-shot | mIoU: 17.43 -> 22.52 (+5.09) | Improved |
| pre-train on YouTube-Occ with 10% labels | Symphonies · Occ-ScanNet few-shot | mIoU: 34.18 -> 36.54 (+2.36) | Improved |
| scale pre-training data from 50K subset to full YouTube-Occ (ablation) | Symphonies · NYUv2 full labels | mIoU: 30.87 -> 31.45 | Improved |
| add region loss (ablation) | Symphonies pre-training · NYUv2 | mIoU: 29.70 -> 31.12 | Improved |
Closest papers by meaning
- LowRankOcc: Tensor Decomposition and Low-Rank Recovery for Vision-based 3D Semantic Occupancy Prediction (CVPR 2024 · similarity 0.88)
- OccMamba: Semantic Occupancy Prediction with State Space Models (CVPR 2025 · similarity 0.88)
- OctOcc: High-Resolution 3D Occupancy Prediction with Octree (AAAI 2024 · similarity 0.88)
- AdaOcc: Adaptive 3D Occupancy Prediction for Embodied Tasks (NeurIPS 2026 · similarity 0.87)
- CSV-Occ: Fusing Multi-frame Alignment for Occupancy Prediction with Temporal Cross State Space Model and Central Voting Mechanism (ICML 2025 · similarity 0.87)
- 3D Occupancy Prediction with Low-Resolution Queries via Prototype-aware View Transformation (CVPR 2025 · 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.
