Research map / Point clouds / Registration and matching
EP2P-Loc: End-to-End 3D Point to 2D Pixel Localization for Large-Scale Visual Localization
Minjung Kim, Junseo Koo, Gunhee Kim
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 |
|---|---|---|---|
| use EP2P-Loc instead of HLoc and other image-to-image methods | visual localization · 2D-3D-S 3 folds | localization recall: Fold 1 (0.1,1.0): 84.14 -> 87.59 | Improved |
| use EP2P-Loc instead of DeepI2P | image-to-point cloud registration · KITTI | RTE (m): 3.28 -> 1.32 | Improved |
| add Invisible 3D Point Removal (IPR) (ablation) | EP2P-Loc · 2D-3D-S | runtime (sec): 6.981 -> 3.482 | Improved |
| add Invisible 3D Point Removal (IPR) (ablation) | EP2P-Loc · 2D-3D-S localization recall | recall: Fold 1 (0.1,1.0): 88.12 -> 87.59; Fold 3: 83.23 -> 85.24 | Mixed |
| use differentiable PnP instead of EPnP with RANSAC (ablation) | EP2P-Loc · 2D-3D-S Fold 1 (0.1,1.0) | recall: 85.34 -> 87.59 | Improved |
| use differentiable PnP instead of EPnP with RANSAC (ablation) | EP2P-Loc · 2D-3D-S | runtime (sec): 12.53 -> 3.482 | Improved |
| use depth maps with RGB-D-to-RGB-D DSAC* instead of image-to-image DSAC* | DSAC* · 2D-3D-S | localization recall: Fold 3 (1.0,5.0): 19.84 -> 13.54 | Hurt |
Closest papers by meaning
- P2-Net: Joint Description and Detection of Local Features for Pixel and Point Matching (ICCV 2021 · similarity 0.88)
- Yes, we CANN: Constrained Approximate Nearest Neighbors for Local Feature-Based Visual Localization (ICCV 2023 · similarity 0.88)
- SG2Loc: Sequential Visual Localization on 3D Scene Graphs (ICML 2026 · similarity 0.88)
- Learning to Produce Semi-dense Correspondences for Visual Localization (CVPR 2024 · similarity 0.87)
- 360Loc: A Dataset and Benchmark for Omnidirectional Visual Localization with Cross-device Queries (CVPR 2024 · similarity 0.86)
- Rooms from Motion: Un-posed Indoor 3D Object Detection as Localization and Mapping (NeurIPS 2025 · similarity 0.86)
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.
