Research map / 3D vision / 3D generation
Learning Spatial-Temporal Consistency for 3D Semantic Scene Completion
Yujie Xue, Meng Wang, Ruihui Li, Fan Wu, Zhizhong Liu, Zhuo Tang, Kenli Li
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 Coarse Occupancy Generation and Unified Spatiotemporal Refinement and TSA to the deformable cross-attention baseline (ablation) | ConSSC · SemanticKITTI val | mIoU (%): 16.95 -> 19.10 | Improved |
| add the soft mask in Unified Spatiotemporal Refinement (ablation) | ConSSC · SemanticKITTI val | mIoU (%): 18.37 -> 19.10 | Improved |
| use similarity-guided USR instead of convolutional fusion (ablation) | ConSSC · SemanticKITTI val | mIoU (%): 18.28 -> 19.10 | Improved |
| use SAM masks instead of image features in TSA (ablation) | ConSSC · SemanticKITTI val | mIoU (%): 18.69 -> 19.10 mIoU; IoU 49.19 -> 49.10 | Mixed |
| add the temporal attention layer in TSA (ablation) | ConSSC · SemanticKITTI val | mIoU (%): 18.87 -> 19.10 | Improved |
| increase historical frames from one to four (ablation) | ConSSC · SemanticKITTI val | mIoU (%): 18.54 -> 19.10 | Improved |
| add a fifth historical frame beyond four (ablation) | ConSSC · SemanticKITTI val | mIoU (%): 19.10 -> 18.99 | No clear effect |
| use temporal multi-frame input instead of single-frame | ConSSC vs ScanSSC and CGFormer · SemanticKITTI hidden test | IoU (%): exceeds ScanSSC by 3.63% IoU | 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.
- SOAP — SOAP: Vision-Centric 3D Semantic Scene Completion with Scene-Adaptive Decoder and Occluded Region-Aware View Projection (CVPR 2025) · better on 4 of 4 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
- Learning Temporal 3D Semantic Scene Completion via Optical Flow Guidance (NeurIPS 2025 · similarity 0.91)
- Bi-SSC: Geometric-Semantic Bidirectional Fusion for Camera-based 3D Semantic Scene Completion (CVPR 2024 · similarity 0.90)
- Context and Geometry Aware Voxel Transformer for Semantic Scene Completion (NeurIPS 2024 · similarity 0.89)
- Memory-Augmented Re-Completion for 3D Semantic Scene Completion (AAAI 2025 · similarity 0.89)
- VLScene: Vision-Language Guidance Distillation for Camera-Based 3D Semantic Scene Completion (AAAI 2025 · similarity 0.88)
- H2GFormer: Horizontal-to-Global Voxel Transformer for 3D Semantic Scene Completion (AAAI 2024 · similarity 0.88)
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.
