Research map / Detection and segmentation / 3D detection and driving
CoDA: Collaborative Novel Box Discovery and Cross-modal Alignment for Open-vocabulary 3D Object Detection
Yang Cao, Yihan Zeng, Hang Xu, Dan Xu
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 3D-NOD novel object discovery to distillation base (ablation) | Distillation 3DETR · SUN-RGBD | AP_Novel: 3.28 -> 5.48 | Improved |
| add 3D-NOD to distillation base (ablation) | Distillation 3DETR · SUN-RGBD | AR_Novel: 19.87 -> 33.45 | Improved |
| add 3D-NOD to distillation base (ablation) | Distillation 3DETR · SUN-RGBD | AP_Base: 33.00 -> 32.07 | Hurt |
| plain text-point alignment instead of DCMA (ablation) | 3D-NOD + Distillation · SUN-RGBD | AP_Novel: 5.48 -> 0.85 | Hurt |
| plain text-point alignment instead of DCMA (ablation) | 3D-NOD + Distillation · SUN-RGBD | AP_Base: 32.07 -> 35.23 | Improved |
| use discovery-driven DCMA instead of plain alignment (ablation) | 3D-NOD + Distillation · SUN-RGBD | AP_Novel: 0.85 -> 6.71 | Improved |
| add DCMA to 3D-NOD on 3DETR+CLIP (ablation) | 3DETR + CLIP + 3D-NOD · SUN-RGBD | AP_Base: 26.08 -> 38.72 | Improved |
| add DCMA to 3D-NOD on 3DETR+CLIP (ablation) | 3DETR + CLIP + 3D-NOD · SUN-RGBD | AP_Novel: 5.30 -> 6.71 | Improved |
| add 3D-NOD to 3DETR+CLIP (ablation) | 3DETR + CLIP · SUN-RGBD | AP_Base: 30.56 -> 26.08 | Hurt |
| class-agnostic distillation instead of 3DETR+CLIP (ablation) | 3DETR · SUN-RGBD | AP_Novel: 3.61 -> 3.28 | No clear effect |
| disable semantic and geometry priors (thresholds 0.0) (ablation) | 3D-NOD · SUN-RGBD | AP_Novel: 3.28 -> 6.71 | Improved |
| raise thresholds from 0.3/0.3 to 0.5/0.3 (ablation) | 3D-NOD · SUN-RGBD | AP_Novel: 6.71 -> 5.70 | Hurt |
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.
- Det-CLIP^2 — DetCLIP: Dictionary-Enriched Visual-Concept Paralleled Pre-training for Open-world Detection (NeurIPS 2022) · better on 5 of 5 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
- Hierarchical Cross-Modal Alignment for Open-Vocabulary 3D Object Detection (AAAI 2025 · similarity 0.92)
- ImOV3D: Learning Open Vocabulary Point Clouds 3D Object Detection from Only 2D Images (NeurIPS 2024 · similarity 0.89)
- FM-OV3D: Foundation Model-Based Cross-Modal Knowledge Blending for Open-Vocabulary 3D Detection (AAAI 2024 · similarity 0.88)
- Cross-Modal Distillation for 2D/3D Multi-Object Discovery from 2D Motion (CVPR 2025 · similarity 0.88)
- Open3DIS: Open-Vocabulary 3D Instance Segmentation with 2D Mask Guidance (CVPR 2024 · similarity 0.88)
- LabelAny3D: Label Any Object 3D in the Wild (NeurIPS 2025 · 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.
