pith. sign in

arxiv: 2606.08205 · v1 · pith:JP2RC3OHnew · submitted 2026-06-06 · 💻 cs.CV

Empowering Feed-Forward Reconstruction Models with Metric Scale via Satellite Images

Xianghui Ze , Yongjian Luo , Mengjun Chao , Zhenbo Song , Jianfeng Lu , Yujiao Shi This is my paper

Pith reviewed 2026-06-27 19:40 UTC · model grok-4.3

classification 💻 cs.CV
keywords 3D reconstructionmetric scalesatellite imageryfeed-forward modelsscale ambiguitycamera pose estimationcross-view consistencydepth estimation
0
0 comments X

The pith

Satellite imagery supplies the missing global metric scale to feed-forward 3D reconstruction models.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Feed-forward 3D reconstruction models recover scene geometry only up to an unknown scale, which blocks metric applications. The paper shows that readily available satellite images can serve as an external global reference to remove this ambiguity. Given a coarse camera pose, the method pulls the matching local satellite patch and fuses it with the reconstruction backbone through bidirectional cross-view interaction. Consistency between the reconstructed scene and the satellite patch then determines absolute scale, refines geometry, and yields metric camera poses. The approach improves metric depth, point-cloud reconstruction, and localization on KITTI, nuScenes, and Oxford RobotCar while retaining cross-dataset generalization.

Core claim

By retrieving a local satellite patch from a coarse pose and enforcing consistency with the feed-forward reconstruction output through bidirectional cross-view interaction, the model infers absolute scale, refines scene geometry, and produces camera poses in a metric coordinate frame without needing large-scale metric annotations or precise calibration.

What carries the argument

Bidirectional cross-view interaction between the feed-forward reconstruction backbone and the retrieved satellite patch that enforces geometric consistency to resolve scale.

If this is right

  • Metric depth estimation improves on KITTI, nuScenes, and Oxford RobotCar.
  • Multi-view point-cloud reconstruction gains accuracy in a metric frame.
  • Cross-view camera localization performance increases while keeping strong generalization.
  • The same backbone can now produce metric outputs across different datasets and regions without retraining on metric labels.
  • Applications that need absolute distances become feasible without costly calibration or annotation pipelines.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The framework could be tested with other overhead references such as aerial or drone imagery when satellite coverage is sparse.
  • In rapidly changing urban areas the method may need periodic satellite updates to avoid drift from outdated geometry.
  • Combining the satellite consistency loss with additional sensors could further stabilize scale in low-texture or night scenes.

Load-bearing premise

The method needs a coarse camera pose to fetch the right satellite patch and assumes the satellite image supplies a reliable, up-to-date match to the ground-level scene geometry.

What would settle it

Run the model on a scene where the retrieved satellite patch is deliberately taken from a different geographic location or an outdated capture and check whether metric scale, depth accuracy, or pose estimates collapse relative to ground truth.

Figures

Figures reproduced from arXiv: 2606.08205 by Jianfeng Lu, Mengjun Chao, Xianghui Ze, Yongjian Luo, Yujiao Shi, Zhenbo Song.

Figure 1
Figure 1. Figure 1: By incorporating satellite patches retrieved from coarse GPS signals, our method [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overview of the proposed framework. Given a coarse ground-view pose, our method [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Visualization of monocular metric depth estimation. The top-left shows the ground [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Visualization of multi-view point-cloud estimation. Our method produces reconstruc [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Visualization of localization results on the KITTI and Oxford RobotCar datasets. In [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗
read the original abstract

Feed-forward 3D reconstruction models have recently shown strong generalization across diverse scenes, yet most of them recover geometry only up to an unknown global scale. This scale ambiguity limits their use in applications that require metric understanding of the environment. Existing metric reconstruction methods commonly rely on large-scale metric annotations or accurate camera calibration, both of which are costly or unreliable in many real-world settings. We propose a satellite-guided framework for resolving scale ambiguity in feed-forward 3D reconstruction. The key idea is to use readily available satellite imagery as a global metric reference. Given a coarse camera pose, our method retrieves a local satellite patch and integrates it with a feed-forward reconstruction backbone through bidirectional cross-view interaction. By enforcing consistency between the reconstructed scene and the satellite reference, the model infers absolute scale, refines scene geometry, and estimates camera pose in a metric coordinate frame. Experiments on KITTI, nuScenes, and Oxford RobotCar show consistent improvements in metric depth estimation, multi-view point-cloud reconstruction, and cross-view camera localization, while preserving strong generalization across datasets and geographic regions.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper proposes a satellite-guided framework to resolve scale ambiguity in feed-forward 3D reconstruction models. Given a coarse camera pose, it retrieves a local satellite patch as a global metric reference and integrates it with a feed-forward backbone via bidirectional cross-view interaction. Consistency enforcement between the reconstructed scene and satellite data is used to infer absolute scale, refine geometry, and estimate metric camera poses. Experiments on KITTI, nuScenes, and Oxford RobotCar report improvements in metric depth estimation, multi-view point-cloud reconstruction, and cross-view localization while preserving generalization.

Significance. If the consistency mechanism functions as described, the approach offers a practical route to metric-scale reconstruction that avoids costly metric annotations or precise calibration by leveraging ubiquitous satellite imagery, which could extend feed-forward models to applications requiring absolute scale such as robotics and mapping.

major comments (2)
  1. [Abstract and §3] Abstract and §3 (method overview): the central claim that consistency enforcement recovers absolute scale rests on retrieving a matching satellite patch, yet the manuscript provides no quantitative analysis of the required coarse-pose accuracy, sensitivity to pose error, or fallback when overlap fails due to temporal change or occlusion; this assumption is load-bearing because mismatch nullifies the metric reference signal.
  2. [§4] §4 (experiments): while improvements on KITTI, nuScenes, and Oxford RobotCar are reported, the evaluation does not include controlled tests of pose perturbation or geographic mismatch between ground and satellite views, leaving the robustness of the metric-scale claim unverified.
minor comments (2)
  1. [§3] Notation for the bidirectional cross-view interaction module should be defined explicitly with equations rather than prose description only.
  2. [Figures] Figure captions should state the exact satellite source and resolution used for each dataset to allow reproduction.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments, which highlight important aspects of robustness in our satellite-guided approach. We address each major comment below and commit to revisions that strengthen the manuscript.

read point-by-point responses

  1. Referee: [Abstract and §3] Abstract and §3 (method overview): the central claim that consistency enforcement recovers absolute scale rests on retrieving a matching satellite patch, yet the manuscript provides no quantitative analysis of the required coarse-pose accuracy, sensitivity to pose error, or fallback when overlap fails due to temporal change or occlusion; this assumption is load-bearing because mismatch nullifies the metric reference signal.

    Authors: We agree that the central claim depends on successful satellite patch retrieval and that the manuscript lacks explicit quantitative analysis of coarse-pose accuracy requirements, sensitivity to errors, and handling of mismatches (e.g., temporal change or occlusion). While the method description notes the use of a coarse pose and the experiments demonstrate gains on the evaluated datasets, we will add a dedicated sensitivity study in the revision. This will include controlled pose perturbations, performance curves under increasing error, and discussion of fallback mechanisms when overlap is insufficient. revision: yes

  2. Referee: [§4] §4 (experiments): while improvements on KITTI, nuScenes, and Oxford RobotCar are reported, the evaluation does not include controlled tests of pose perturbation or geographic mismatch between ground and satellite views, leaving the robustness of the metric-scale claim unverified.

    Authors: We acknowledge that the current experimental section reports improvements on the three datasets but does not include controlled tests for pose perturbation or geographic mismatch. These additional evaluations would better substantiate the robustness of the metric-scale recovery. We will incorporate such controlled experiments in the revised manuscript, including synthetic pose noise injection and tests across geographic regions with varying satellite-ground alignment. revision: yes

Circularity Check

0 steps flagged

No circularity; method relies on external satellite reference rather than internal fits or self-citations

full rationale

The paper proposes integrating external satellite imagery as a global metric reference with a feed-forward reconstruction backbone via bidirectional cross-view interaction and consistency enforcement. No equations, parameter fittings to data subsets, or self-citation chains are described in the provided text that would reduce the claimed metric scale inference to a tautology or renamed input. The approach is self-contained against external benchmarks (satellite data), with the central claim depending on the validity of the external reference rather than any self-definitional or fitted-input reduction. This is the normal honest finding for papers without visible internal circular steps.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are stated. The central claim rests on the unstated assumption that satellite patches can be reliably aligned with ground-level reconstructions.

pith-pipeline@v0.9.1-grok · 5730 in / 1011 out tokens · 20605 ms · 2026-06-27T19:40:15.068429+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

45 extracted references · 10 canonical work pages · 7 internal anchors

  1. [1]

    Accurate 3-d reconstruc- tion under iot environments and its applications to augmented reality.IEEE Transactions on Industrial Informatics, 17(3):2090–2100, 2020

    Mingwei Cao, Liping Zheng, Wei Jia, Huimin Lu, and Xiaoping Liu. Accurate 3-d reconstruc- tion under iot environments and its applications to augmented reality.IEEE Transactions on Industrial Informatics, 17(3):2090–2100, 2020

    2090

  2. [2]

    Learning-based 3d reconstruction in autonomous driving: A comprehensive survey.IEEE Transactions on Intelligent Transportation Systems, 2025

    Liewen Liao, Weihao Yan, Wang Xu, Ming Yang, Songan Zhang, and Hongtei Eric Tseng. Learning-based 3d reconstruction in autonomous driving: A comprehensive survey.IEEE Transactions on Intelligent Transportation Systems, 2025. 10

    2025

  3. [3]

    A survey of embodied ai: From simulators to research tasks.IEEE Transactions on Emerging Topics in Computational Intelligence, 6(2):230–244, 2022

    Jiafei Duan, Samson Yu, Hui Li Tan, Hongyuan Zhu, and Cheston Tan. A survey of embodied ai: From simulators to research tasks.IEEE Transactions on Emerging Topics in Computational Intelligence, 6(2):230–244, 2022

    2022

  4. [4]

    A survey of structure from motion*.Acta Numerica, 26:305–364, 2017

    Onur ¨Ozye¸ sil, Vladislav Voroninski, Ronen Basri, and Amit Singer. A survey of structure from motion*.Acta Numerica, 26:305–364, 2017

    2017

  5. [5]

    Multi-view stereo: A tutorial.Foundations and Trends in Computer Graphics and Vision, 9(1-2):1–148, 2015

    Yasutaka Furukawa and Carlos Hern´ andez. Multi-view stereo: A tutorial.Foundations and Trends in Computer Graphics and Vision, 9(1-2):1–148, 2015

    2015

  6. [6]

    Mast3r-slam: Real-time dense slam with 3d reconstruction priors

    Riku Murai, Eric Dexheimer, and Andrew J Davison. Mast3r-slam: Real-time dense slam with 3d reconstruction priors. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 16695–16705, 2025

    2025

  7. [7]

    Vggt: Visual geometry grounded transformer

    Jianyuan Wang, Minghao Chen, Nikita Karaev, Andrea Vedaldi, Christian Rupprecht, and David Novotny. Vggt: Visual geometry grounded transformer. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 5294–5306, 2025

    2025

  8. [8]

    $\pi^3$: Permutation-Equivariant Visual Geometry Learning

    Yifan Wang, Jianjun Zhou, Haoyi Zhu, Wenzheng Chang, Yang Zhou, Zizun Li, Junyi Chen, Jiangmiao Pang, Chunhua Shen, and Tong He. π3: Permutation-equivariant visual geometry learning.arXiv preprint arXiv:2507.13347, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  9. [9]

    Unidepthv2: Universal monocular metric depth estimation made simpler

    Luigi Piccinelli, Christos Sakaridis, Yung-Hsu Yang, Mattia Segu, Siyuan Li, Wim Abbeloos, and Luc Van Gool. Unidepthv2: Universal monocular metric depth estimation made simpler. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    2025

  10. [10]

    MoGe-2: Accurate Monocular Geometry with Metric Scale and Sharp Details

    Ruicheng Wang, Sicheng Xu, Yue Dong, Yu Deng, Jianfeng Xiang, Zelong Lv, Guangzhong Sun, Xin Tong, and Jiaolong Yang. Moge-2: Accurate monocular geometry with metric scale and sharp details.arXiv preprint arXiv:2507.02546, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  11. [11]

    Depth Anything 3: Recovering the Visual Space from Any Views

    Haotong Lin, Sili Chen, Junhao Liew, Donny Y Chen, Zhenyu Li, Guang Shi, Jiashi Feng, and Bingyi Kang. Depth anything 3: Recovering the visual space from any views.arXiv preprint arXiv:2511.10647, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  12. [12]

    Metricdepth: Enhancing monocular depth estimation with deep metric learning.arXiv preprint arXiv:2412.20390, 2024

    Chunpu Liu, Guanglei Yang, Wangmeng Zuo, and Tianyi Zan. Metricdepth: Enhancing monocular depth estimation with deep metric learning.arXiv preprint arXiv:2412.20390, 2024

    work page arXiv 2024

  13. [13]

    Resolving scale ambiguity in multi-view 3d reconstruction using dual-pixel sensors

    Kohei Ashida, Hiroaki Santo, Fumio Okura, and Yasuyuki Matsushita. Resolving scale ambiguity in multi-view 3d reconstruction using dual-pixel sensors. InEuropean Conference on Computer Vision, pages 162–178. Springer, 2024

    2024

  14. [14]

    Convolutional cross-view pose estimation

    Zimin Xia, Olaf Booij, and Julian FP Kooij. Convolutional cross-view pose estimation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(5):3813–3831, 2023

    2023

  15. [15]

    Pixelwise view selection for unstructured multi-view stereo

    Johannes L Sch¨ onberger, Enliang Zheng, Jan-Michael Frahm, and Marc Pollefeys. Pixelwise view selection for unstructured multi-view stereo. InEuropean conference on computer vision, pages 501–518. Springer, 2016

    2016

  16. [16]

    Structure from motion photogrammetry in forestry: A review.Current Forestry Reports, 5(3):155–168, 2019

    Jakob Iglhaut, Carlos Cabo, Stefano Puliti, Livia Piermattei, James OConnor, and Jacque- line Rosette. Structure from motion photogrammetry in forestry: A review.Current Forestry Reports, 5(3):155–168, 2019

    2019

  17. [17]

    Cambridge university press, 2003

    Richard Hartley and Andrew Zisserman.Multiple view geometry in computer vision. Cambridge university press, 2003. 11

    2003

  18. [18]

    Multi-view stereo revisited

    Michael Goesele, Brian Curless, and Steven M Seitz. Multi-view stereo revisited. In 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’06), volume 2, pages 2402–2409. IEEE, 2006

    2006

  19. [19]

    Dust3r: Geometric 3d vision made easy

    Shuzhe Wang, Vincent Leroy, Yohann Cabon, Boris Chidlovskii, and Jerome Revaud. Dust3r: Geometric 3d vision made easy. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 20697–20709, 2024

    2024

  20. [20]

    Grounding image matching in 3d with mast3r

    Vincent Leroy, Yohann Cabon, and J´ erˆ ome Revaud. Grounding image matching in 3d with mast3r. InEuropean conference on computer vision, pages 71–91. Springer, 2024

    2024

  21. [21]

    Splatt3R: Zero-shot Gaussian Splatting from Uncalibrated Image Pairs

    Brandon Smart, Chuanxia Zheng, Iro Laina, and Victor Adrian Prisacariu. Splatt3r: Zero- shot gaussian splatting from uncalibrated image pairs.arXiv preprint arXiv:2408.13912, 2024

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  22. [22]

    3d reconstruction with spatial memory

    Hengyi Wang and Lourdes Agapito. 3d reconstruction with spatial memory. In2025 International Conference on 3D Vision (3DV), pages 78–89. IEEE, 2025

    2025

  23. [23]

    Fast3r: Towards 3d reconstruction of 1000+ images in one forward pass

    Jianing Yang, Alexander Sax, Kevin J Liang, Mikael Henaff, Hao Tang, Ang Cao, Joyce Chai, Franziska Meier, and Matt Feiszli. Fast3r: Towards 3d reconstruction of 1000+ images in one forward pass. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 21924–21935, 2025

    2025

  24. [24]

    Flare: Feed-forward geometry, appearance and camera estimation from uncalibrated sparse views

    Shangzhan Zhang, Jianyuan Wang, Yinghao Xu, Nan Xue, Christian Rupprecht, Xiaowei Zhou, Yujun Shen, and Gordon Wetzstein. Flare: Feed-forward geometry, appearance and camera estimation from uncalibrated sparse views. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 21936–21947, 2025

    2025

  25. [25]

    Block-Sparse Global Attention for Efficient Multi-View Geometry Transformers

    Chung-Shien Brian Wang, Christian Schmidt, Jens Piekenbrinck, and Bastian Leibe. Faster vggt with block-sparse global attention.arXiv preprint arXiv:2509.07120, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  26. [26]

    Vigor: Cross-view image geo-localization beyond one-to-one retrieval

    Sijie Zhu, Taojiannan Yang, and Chen Chen. Vigor: Cross-view image geo-localization beyond one-to-one retrieval. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 3640–3649, 2021

    2021

  27. [27]

    Geoclip: Clip-inspired alignment between locations and images for effective worldwide geo-localization.Advances in Neural Information Processing Systems, 36:8690–8701, 2023

    Vicente Vivanco Cepeda, Gaurav Kumar Nayak, and Mubarak Shah. Geoclip: Clip-inspired alignment between locations and images for effective worldwide geo-localization.Advances in Neural Information Processing Systems, 36:8690–8701, 2023

    2023

  28. [28]

    Gama: Cross-view video geo-localization

    Shruti Vyas, Chen Chen, and Mubarak Shah. Gama: Cross-view video geo-localization. In European Conference on Computer Vision, pages 440–456. Springer, 2022

    2022

  29. [29]

    Cvlnet: Cross-view semantic corre- spondence learning for video-based camera localization

    Yujiao Shi, Xin Yu, Shan Wang, and Hongdong Li. Cvlnet: Cross-view semantic corre- spondence learning for video-based camera localization. InAsian Conference on Computer Vision, pages 123–141. Springer, 2022

    2022

  30. [30]

    Where am i looking at? joint location and orientation estimation by cross-view matching

    Yujiao Shi, Xin Yu, Dylan Campbell, and Hongdong Li. Where am i looking at? joint location and orientation estimation by cross-view matching. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4064–4072, 2020

    2020

  31. [31]

    Beyond cross-view image retrieval: Highly accurate vehicle localization using satellite image

    Yujiao Shi and Hongdong Li. Beyond cross-view image retrieval: Highly accurate vehicle localization using satellite image. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17010–17020, 2022

    2022

  32. [32]

    Boosting 3-dof ground-to-satellite camera localization accuracy via geometry-guided cross-view transformer

    Yujiao Shi, Fei Wu, Akhil Perincherry, Ankit Vora, and Hongdong Li. Boosting 3-dof ground-to-satellite camera localization accuracy via geometry-guided cross-view transformer. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 21516–21526, 2023. 12

    2023

  33. [33]

    Slicematch: Geometry-guided aggregation for cross-view pose estimation

    Ted Lentsch, Zimin Xia, Holger Caesar, and Julian FP Kooij. Slicematch: Geometry-guided aggregation for cross-view pose estimation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17225–17234, 2023

    2023

  34. [34]

    Geodistill: Geometry-guided self-distillation for weakly supervised cross-view localization

    Shaowen Tong, Zimin Xia, Alexandre Alahi, Xuming He, and Yujiao Shi. Geodistill: Geometry-guided self-distillation for weakly supervised cross-view localization. InProceed- ings of the IEEE/CVF International Conference on Computer Vision, pages 25357–25366, 2025

    2025

  35. [35]

    Learning dense flow field for highly-accurate cross-view camera localization.Advances in Neural Information Processing Systems, 36:70612–70625, 2023

    Zhenbo Song, Jianfeng Lu, Yujiao Shi, et al. Learning dense flow field for highly-accurate cross-view camera localization.Advances in Neural Information Processing Systems, 36:70612–70625, 2023

    2023

  36. [36]

    Fine-grained cross- view geo-localization using a correlation-aware homography estimator.Advances in Neural Information Processing Systems, 36:5301–5319, 2023

    Xiaolong Wang, Runsen Xu, Zhuofan Cui, Zeyu Wan, and Yu Zhang. Fine-grained cross- view geo-localization using a correlation-aware homography estimator.Advances in Neural Information Processing Systems, 36:5301–5319, 2023

    2023

  37. [37]

    F G2: Fine-grained cross-view localization by fine-grained feature matching

    Zimin Xia and Alexandre Alahi. F G2: Fine-grained cross-view localization by fine-grained feature matching. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 6362–6372, 2025

    2025

  38. [38]

    Depth Pro: Sharp Monocular Metric Depth in Less Than a Second

    Aleksei Bochkovskii, Ama˜ aG ¸l Delaunoy, Hugo Germain, Marcel Santos, Yichao Zhou, Stephan R Richter, and Vladlen Koltun. Depth pro: Sharp monocular metric depth in less than a second.arXiv preprint arXiv:2410.02073, 2024

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  39. [39]

    Grounding image matching in 3d with mast3r, 2024

    Vincent Leroy, Yohann Cabon, and Jerome Revaud. Grounding image matching in 3d with mast3r, 2024

    2024

  40. [40]

    Vision meets robotics: The kitti dataset.The International Journal of Robotics Research, 32(11):1231–1237, 2013

    Andreas Geiger, Philip Lenz, Christoph Stiller, and Raquel Urtasun. Vision meets robotics: The kitti dataset.The International Journal of Robotics Research, 32(11):1231–1237, 2013

    2013

  41. [41]

    nuScenes: A multimodal dataset for autonomous driving

    Holger Caesar, Varun Bankiti, Alex H. Lang, Sourabh Vora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Giancarlo Baldan, and Oscar Beijbom. nuscenes: A multimodal dataset for autonomous driving.arXiv preprint arXiv:1903.11027, 2019

    work page internal anchor Pith review Pith/arXiv arXiv 1903

  42. [42]

    Panoptic nuscenes: A large-scale benchmark for lidar panoptic segmentation and tracking.arXiv preprint arXiv:2109.03805, 2021

    Whye Kit Fong, Rohit Mohan, Juana Valeria Hurtado, Lubing Zhou, Holger Caesar, Oscar Beijbom, and Abhinav Valada. Panoptic nuscenes: A large-scale benchmark for lidar panoptic segmentation and tracking.arXiv preprint arXiv:2109.03805, 2021

    work page arXiv 2021

  43. [43]

    https://developers.google.com/maps/documentation/mapsstatic/intro

  44. [44]

    Bevsplat: Resolving height ambiguity via feature-based gaussian primitives for weakly-supervised cross-view localization.arXiv preprint arXiv:2502.09080, 2025

    Qiwei Wang, Shaoxun Wu, and Yujiao Shi. Bevsplat: Resolving height ambiguity via feature-based gaussian primitives for weakly-supervised cross-view localization.arXiv preprint arXiv:2502.09080, 2025

    work page arXiv 2025

  45. [45]

    Visual cross-view metric localization with dense uncertainty estimates

    Zimin Xia, Olaf Booij, Marco Manfredi, and Julian FP Kooij. Visual cross-view metric localization with dense uncertainty estimates. InComputer Vision–ECCV 2022: 17th European Conference, Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part XXXIX, pages 90–106. Springer, 2022. 13

    2022