Research map / Point clouds / Point cloud classification
3D-SPS: Single-Stage 3D Visual Grounding via Referred Point Progressive Selection
Junyu Luo, Jiahui Fu, Xianghao Kong, Chen Gao, Haibing Ren, Hao Shen, Huaxia Xia, Si Liu
Check 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 |
|---|---|---|---|
| Use single-stage progressive referred-point selection | 3D-SPS · ScanRefer Overall, 3D only | Acc@0.5: 34.47 -> 36.43 | Improved |
| Use single-stage progressive referred-point selection | 3D-SPS · ScanRefer Overall, 2D + 3D | Acc@0.5: 34.67 -> 36.98 | Improved |
| Use TPM progressive selection | 3D-SPS · Nr3D Overall | accuracy: +2.3% | Improved |
| Use TPM progressive selection | 3D-SPS · Sr3D Overall | accuracy: +4.7% | Improved |
| Add description relevance score s_d to DKS (ablation) | 3D-SPS on ScanRefer · ScanRefer Overall, 3D only | Acc@0.5: 34.95 -> 36.43 | Improved |
| Add object confidence score s_o to DKS (ablation) | 3D-SPS on ScanRefer · ScanRefer Overall, 3D only | Acc@0.5: 35.19 -> 36.43 | Improved |
| Replace FPS with DKS (ablation) | 3D-SPS on ScanRefer · ScanRefer Overall, 3D only | Acc@0.5: 31.88 -> 36.43 | Improved |
| Use 4 TPM layers rather than 1, 2, or 3 (ablation) | 3D-SPS on ScanRefer · ScanRefer Overall, 3D only | Acc@0.5: 33.13, 33.97, 34.53 -> 36.43 | Improved |
| Increase TPM depth from 4 to 5 layers (ablation) | 3D-SPS on ScanRefer · ScanRefer Overall, 3D only | Acc@0.5: 36.43 -> 36.07 | Hurt |
| Add progressive keypoint selection, reducing 512 keypoints to 32 (ablation) | 3D-SPS on ScanRefer · ScanRefer Overall, 3D only | Acc@0.5: 34.98 -> 36.43 | Improved |
| Use 3D-SPS on Sr3D Easy rather than LanguageRefer | 3D-SPS · Sr3D Easy | accuracy: 58.9% -> 56.2% ± 0.6% | Hurt |
Later papers that compared with this method
3 later accepted papers report 3D-SPS as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- WildRefer: 3D Object Localization in Large-scale Dynamic Scenes with Multi-modal Visual Data and Natural Language (ECCV 2024 · better on 1 of 1 shared result)
- EDA: Explicit Text-Decoupling and Dense Alignment for 3D Visual Grounding (CVPR 2023 · better on 1 of 1 shared result)
- Language Conditioned Spatial Relation Reasoning for 3D Object Grounding (NeurIPS 2022 · better on 2 of 2 shared results)
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.
- 3DVG-Transformer — 3DVG-Transformer: Relation Modeling for Visual Grounding on Point Clouds (ICCV 2021) · better on 2 of 2 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
- Multi-Object 3D Grounding with Dynamic Modules and Language-Informed Spatial Attention (NeurIPS 2024 · similarity 0.89)
- From Objects to Anywhere: A Holistic Benchmark for Multi-level Visual Grounding in 3D Scenes (NeurIPS 2025 · similarity 0.88)
- LIBA: Language Instructed Multi-granularity Bridge Assistant for 3D Visual Grounding (AAAI 2025 · similarity 0.88)
- SPAZER: Spatial-Semantic Progressive Reasoning Agent for Zero-shot 3D Visual Grounding (NeurIPS 2025 · similarity 0.88)
- G^3-LQ: Marrying Hyperbolic Alignment with Explicit Semantic-Geometric Modeling for 3D Visual Grounding (CVPR 2024 · similarity 0.88)
- CoT3DRef: Chain-of-Thoughts Data-Efficient 3D Visual Grounding (ICLR 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.
