Research map / 3D vision / 3D generation
GOLDILOCS: GENERAL OBJECT-LEVEL DETECTION AND LABELING OF CHANGES IN SCENES
Almog Friedlander, Ariel Shamir, Ohad Fried
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 stereo 3D reconstruction and novel view synthesis (ablation) | GOLDILOCS · ChangeSim (10% subset) | multi-class mIoU: 24.3 -> 33.5 | Improved |
| add stereo 3D reconstruction (binary) (ablation) | GOLDILOCS · ChangeSim (10% subset) | binary mIoU: 46.3 -> 65.0 | Improved |
| add stereo 3D reconstruction, Removed class (ablation) | GOLDILOCS · ChangeSim (10% subset) | Removed IoU: 11.2 -> 21.1 | Improved |
| add cross-view voting (ablation) | GOLDILOCS · 3DGS-CD Bench | IoU: 59.57 -> 95.44 | Improved |
| add cross-view voting (ablation) | GOLDILOCS · 3DGS-CD Desk | IoU: 64.42 -> 96.67 | Improved |
| add cross-view voting (ablation) | GOLDILOCS · 3DGS-CD Swap | precision: 79.93 -> 98.72 | Improved |
| add cross-view voting (ablation) | GOLDILOCS · 3DGS-CD Sill | IoU: 66.42 -> 95.23 | Improved |
| use lower-ranked reference images in top-1 selection (ablation) | GOLDILOCS · 3DGS-CD Desk | F1 and IoU | Hurt |
| reduce 3D reconstruction optimization to 10 steps (5%) (ablation) | GOLDILOCS · ChangeSim (1% subset) | mIoU | No clear effect |
| zero-shot GOLDILOCS vs in-domain supervised | GOLDILOCS · VL-CMU-CD | F1: 80.0 -> 61.8 | Hurt |
| use RGB-only GOLDILOCS instead of CYWS-3D RGB-D | GOLDILOCS · RC-3D | mAP: 0.50 -> 0.531 | Improved |
| use GOLDILOCS instead of ZSSCD | GOLDILOCS · VL-CMU-CD | F1: 51.6 -> 61.8 | Improved |
Closest papers by meaning
- Towards Generalizable Scene Change Detection (CVPR 2025 · similarity 0.89)
- Multi-View Pose-Agnostic Change Localization with Zero Labels (CVPR 2025 · similarity 0.89)
- Segment Any Change (NeurIPS 2024 · similarity 0.88)
- Change3D: Revisiting Change Detection and Captioning from A Video Modeling Perspective (CVPR 2025 · similarity 0.88)
- Zero-Shot Scene Change Detection (AAAI 2025 · similarity 0.88)
- RefineAny3D: Depth Refinement as Semantic Alignment for Monocular 3D Detection (NeurIPS 2026 · similarity 0.85)
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.
