Research map / Detection and segmentation / 3D detection and driving
Simultaneous Multi-View Instance Detection With Learned Geometric Soft-Constraints
Ahmed Samy Nassar, Sebastien Lefevre, Jan Dirk Wegner
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 |
|---|---|---|---|
| project detections between views and fuse them (multi-view) instead of monocular detection | SSD512 detector · Pasadena trees | detection mAP: 0.597 -> 0.682 | Improved |
| multi-view fusion instead of monocular detection | SSD512 detector · Mapillary traffic signs | detection mAP: 0.875 -> 0.902 | Improved |
| add camera pose information to the Siamese network (ablation) | FaceNet · Pasadena Re-ID | Re-ID mAP: 0.808 -> 0.842 | Improved |
| add camera pose information to the Siamese network (ablation) | ResNet-50 · Pasadena Re-ID | Re-ID mAP: 0.828 -> 0.863 | Improved |
| add camera pose information to the Siamese network (ablation) | MatchNet · Pasadena Re-ID | Re-ID mAP: 0.843 -> 0.871 | Improved |
| naively combine detections from multiple views without learned projection instead of single-view projection (ablation) | geo-localization · Pasadena trees | MAE (m): 77.41 -> 70.16 | Improved |
| naive multi-view combination instead of single view (ablation) | geo-localization · Mapillary traffic signs | MAE (m): 83.27 -> 64.0 | Improved |
| add Projection Net and Geo Regression Net (full learned model) to multi-view projection (ablation) | geo-localization · Pasadena trees | MAE (m): 70.16 -> 3.13 | Improved |
| add learned Projection Net and Geo Regression Net (ablation) | geo-localization · Mapillary traffic signs | MAE (m): 64.0 -> 4.36 | Improved |
Closest papers by meaning
- Unified Domain Generalization and Adaptation for Multi-View 3D Object Detection (NeurIPS 2024 · similarity 0.87)
- CaMuViD: Calibration-Free Multi-View Detection (CVPR 2025 · similarity 0.85)
- Large Scale Visual Recognition through Adaptation using Joint Representation and Multiple Instance Learning (JMLR 2016 · similarity 0.85)
- Towards Generalizable Multi-Camera 3D Object Detection via Perspective Rendering (AAAI 2025 · similarity 0.85)
- Group3D: MLLM-Guided Semantic Grouping for Open-Vocabulary 3D Object Detection (ECCV 2026 · similarity 0.85)
- Share with Thy Neighbors: Single-View Reconstruction by Cross-Instance Consistency (ECCV 2022 · 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.
