Research map / Detection and segmentation / Weakly supervised detection and segmentation
Diffusion-SS3D: Diffusion Model for Semi-supervised 3D Object Detection
Cheng-Ju Ho, Chen-Hsuan Tai, Yen-Yu Lin, Ming-Hsuan Yang, Yi-Hsuan Tsai
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 |
|---|---|---|---|
| Replace 3DIoUMatch with Diffusion-SS3D | 3D semi-supervised object detection · ScanNet val, 5% labeled | mAP@0.5: 5.4↑ | Improved |
| Replace 3DIoUMatch with Diffusion-SS3D | 3D semi-supervised object detection · SUN RGB-D val, 1% labeled | mAP@0.5: 6.7↑ | Improved |
| Add diffusion denoising for object sizes (ablation) | Diffusion-SS3D · ScanNet 5% labeled | mAP@0.5: 22.8 ± 0.8 -> 26.3 ± 0.2 | Improved |
| Add diffusion denoising for object sizes (ablation) | Diffusion-SS3D · SUN RGB-D 1% labeled | mAP@0.5: 8.0 ± 1.5 -> 13.8 ± 0.7 | Improved |
| Add diffusion denoising for class labels (ablation) | Diffusion-SS3D · ScanNet 5% labeled | mAP@0.5: 22.8 ± 0.8 -> 26.0 ± 0.5 | Improved |
| Add diffusion denoising for class labels (ablation) | Diffusion-SS3D · SUN RGB-D 1% labeled | mAP@0.5: 8.0 ± 1.5 -> 13.3 ± 0.2 | Improved |
| Use both size and label diffusion (ablation) | Diffusion-SS3D · ScanNet 5% labeled | mAP@0.5: 26.3 ± 0.2 -> 27.9 ± 0.3 | Improved |
| Use both size and label diffusion (ablation) | Diffusion-SS3D · SUN RGB-D 1% labeled | mAP@0.5: 13.8 ± 0.7 -> 14.7 ± 1.2 | Improved |
| Add DDIM sampling (ablation) | Diffusion-SS3D teacher pseudo-labeling · ScanNet 5% labeled | mAP@0.5: 22.8 ± 0.8 -> 26.6 ± 0.9 | Improved |
| Add box renewal (ablation) | Diffusion-SS3D teacher pseudo-labeling · SUN RGB-D 1% labeled | mAP@0.5: 8.0 ± 1.5 -> 12.8 ± 0.6 | Improved |
| Increase DDIM sampling from 2 to 4 steps (ablation) | Diffusion-SS3D · SUN RGB-D 1% labeled | mAP@0.5: 14.7 ± 1.2 -> 12.9 ± 0.3 | Hurt |
| Increase DDIM sampling from 2 to 4 steps (ablation) | Diffusion-SS3D · ScanNet 5% labeled | mAP@0.25: 43.5 ± 0.2 -> 44.0 ± 1.1 | Mixed |
Later papers that compared with this method
1 later accepted paper reports Diffusion-SS3D as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Diff3DETR: Agent-based Diffusion Model for Semi-supervised 3D Object Detection (ECCV 2024 · better on 7 of 7 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.
- 3DIoUMatch — 3DIoUMatch: Leveraging IoU Prediction for Semi-Supervised 3D Object Detection (CVPR 2021) · better on 16 of 16 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
- Leveraging Temporal Cues for Semi-Supervised Multi-View 3D Object Detection (CVPR 2025 · similarity 0.90)
- Semi-supervised 3D Object Detection with PatchTeacher and PillarMix (AAAI 2024 · similarity 0.89)
- Dual-Perspective Knowledge Enrichment for Semi-supervised 3D Object Detection (AAAI 2024 · similarity 0.88)
- 3DiffTection: 3D Object Detection with Geometry-Aware Diffusion Features (CVPR 2024 · similarity 0.88)
- DiffuBox: Refining 3D Object Detection with Point Diffusion (NeurIPS 2024 · similarity 0.88)
- Decoupled Pseudo-labeling for Semi-Supervised Monocular 3D Object Detection (CVPR 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.
