Research map / 3D vision / Depth and stereo
Semantic Scene Completion From a Single Depth Image
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 semantic-label supervision to occupancy completion (ablation) | SSCNet · Rendered NYU test set | scene completion IoU: 64.8 -> 70.3 | Helped |
| Add completion-task supervision to visible-surface semantic labeling (ablation) | SSCNet · SUNCG test set, observed surface | semantic segmentation avg. IoU: 51.2% vs. 54.2% | Helped |
| Pre-train on SUNCG synthetic data before fine-tuning on NYU | SSCNet · NYU test set | semantic scene completion avg. IoU: 24.7 -> 30.5 | Helped |
| Replace standard convolutions with 3D dilated convolutions to enlarge the receptive field (ablation) | SSCNet · SUNCG test set | semantic scene completion avg. IoU: 38.0% to 44.3% | Helped |
| Add multi-scale aggregation (ablation) | SSCNet with 3D dilated convolution · SUNCG test set | scene completion IoU: 70.4 -> 73.5 | Helped |
| Add multi-scale aggregation (ablation) | SSCNet with 3D dilated convolution · SUNCG test set | semantic scene completion avg. IoU: 44.3 -> 46.4 | Helped |
| Replace projective TSDF with accurate TSDF (ablation) | SSCNet with multi-scale aggregation · SUNCG test set | semantic scene completion avg. IoU: 33.9 -> 36.3 | Helped |
| Replace accurate TSDF with flipped TSDF (ablation) | SSCNet with multi-scale aggregation · SUNCG test set | semantic scene completion avg. IoU: 36.3 -> 46.4 | Helped |
| Remove empty-voxel sampling for data balancing (ablation) | SSCNet · SUNCG test set | semantic scene completion avg. IoU: 46.4% to 42.7% | Hurt |
Closest papers by meaning
- Not All Voxels Are Equal: Semantic Scene Completion from the Point-Voxel Perspective (AAAI 2022 · similarity 0.92)
- Context and Geometry Aware Voxel Transformer for Semantic Scene Completion (NeurIPS 2024 · similarity 0.92)
- Voxel Proposal Network via Multi-Frame Knowledge Distillation for Semantic Scene Completion (NeurIPS 2024 · similarity 0.91)
- Memory-Augmented Re-Completion for 3D Semantic Scene Completion (AAAI 2025 · similarity 0.90)
- H2GFormer: Horizontal-to-Global Voxel Transformer for 3D Semantic Scene Completion (AAAI 2024 · similarity 0.90)
- Three Cars Approaching within 100m! Enhancing Distant Geometry by Tri-Axis Voxel Scanning for Camera-based Semantic Scene Completion (CVPR 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.