Research map / 3D vision / 3D generation
MV2GF: Multi-view Pedestrian Detection with a Visual Geometric Foundation Model
Taiga Yamane, Satoshi Suzuki, Ryo Masumura, Shota Orihashi, Tomohiro Tanaka, Mana Ihori, Naoki Makishima
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 TGF (ablation) | MVDet baseline · GMVD-D ablation | MODA: 73.4 -> 77.8 | Improved |
| add FPA (ablation) | MVDet baseline · GMVD-D ablation | MODA: 73.4 -> 81.3 | Improved |
| add both TGF and FPA (ablation) | MVDet baseline · GMVD-D ablation | MODA: 73.4 -> 84.8 | Improved |
| add DA3 features to ResNet in TGF (ablation) | MV2GF · GMVD-D ablation | MODA: 81.3 -> 84.8 | Improved |
| add ResNet features to DA3 in TGF (ablation) | MV2GF · GMVD-D ablation | MODA: 77.5 -> 84.8 | Improved |
| use DA3 features from four blocks instead of one (ablation) | MV2GF · GMVD-D ablation | MODA: 84.0 -> 84.8 | Improved |
| use DA3 instead of DINOv2 features (ablation) | MV2GF · GMVD-D ablation | MODA: 82.9 -> 84.8 | Improved |
| use DA3 instead of MapAnything (ablation) | MV2GF · GMVD-D ablation | MODA: 83.0 -> 84.8 | Improved |
| use theta_z 50 cm instead of 12.5 cm (ablation) | FPA · GMVD-D ablation | MODA: 81.6 -> 84.8 | Improved |
| use max pooling instead of mean pooling (ablation) | FPA · GMVD-D ablation | MODA: 84.5 -> 84.8 | Improved |
| add offset prediction (ablation) | MV2GF · GMVD-D ablation | MODP: 79.0 -> 83.1 | Improved |
| add auxiliary per-view loss (ablation) | MV2GF · GMVD-D ablation | MODA: 83.8 -> 84.8 | Improved |
Closest papers by meaning
- Stacked Homography Transformations for Multi-View Pedestrian Detection (ICCV 2021 · similarity 0.87)
- Towards Generalizable Multi-Camera 3D Object Detection via Perspective Rendering (AAAI 2025 · similarity 0.86)
- Multi-View People Detection in Large Scenes via Supervised View-Wise Contribution Weighting (AAAI 2024 · similarity 0.84)
- Multi-View Pedestrian Occupancy Prediction with a Novel Synthetic Dataset (AAAI 2025 · similarity 0.84)
- Unified Domain Generalization and Adaptation for Multi-View 3D Object Detection (NeurIPS 2024 · similarity 0.84)
- GeoBEV: Learning Geometric BEV Representation for Multi-view 3D Object Detection (AAAI 2025 · similarity 0.84)
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.
