Pith. sign in

REVIEW 3 major objections 5 minor 60 references

TrafficLoc: Localizing Traffic Surveillance Cameras in 3D Scenes

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read TrafficLoc claims that a traffic surveillance camera can be localized in a 3D LiDAR map through coarse-to-fine image-to-point-cloud matching, beating prior methods by up to 86% on simulated intersections and transferring to real scenes.

desk verdict Solid new dataset and method for traffic camera localization, but the SOTA claim needs a more controlled evaluation before I'd take it at face value. read the letter →

arxiv 2412.10308 v2 pith:R36IUKOY submitted 2024-12-13 cs.CV

Yan Xia , Yunxiang Lu , Rui Song , Oussema Dhaouadi , João F. Henriques , Daniel Cremers This is my paper · ORCID
classification cs.CV
keywords image-to-pointcloudregistrationtrafficcameralocalizationposeestimationcross-modalfeaturefusionLiDARpointcoarse-to-finematchingsim-to-realtransferCarlaIntersectiondataset
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

TrafficLoc tackles a practical unmet problem: given a photo from a roadside traffic camera and a 3D LiDAR point cloud of the same intersection taken at a different time, compute the camera's 6-DoF pose in that map. The paper argues that existing image-to-point-cloud registration methods degrade sharply at intersections because viewpoints and modalities differ too much, and it introduces a coarse-to-fine method with three training components—a geometry-guided attention loss, an inter-intra contrastive loss, and dense training alignment—to keep cross-modal features focused on geometrically corresponding regions. It also contributes Carla Intersection, a simulated dataset of 75 urban and rural intersections, to train and evaluate the task. On that dataset the method cuts median rotation error from 4.24 degrees to 0.66 degrees and median translation error from 2.82 m to 0.51 m, and it reports new best results on two in-vehicle benchmarks plus qualitative transfer to a real intersection. If these results hold, traffic cameras could be calibrated against existing LiDAR maps automatically, without manual matching or special capture equipment.

What carries the argument

The load-bearing machinery is a coarse-to-fine correspondence pipeline supervised by three purpose-built losses. The Geometry-guided Attention Loss (GAL) takes the raw cross-attention map between image patches and 3D point groups and applies a binary cross-entropy penalty using thresholds on the angular separation for image-to-point attention and on the distance from a point to a camera ray for point-to-image attention, so the fusion module learns to attend only to geometrically plausible regions. The Inter-intra Contrastive Learning (ICL) loss augments standard inter-modal contrastive pairs with negative pairs inside each modality, forcing different image patches and different point groups to remain separable. The Dense Training Alignment (DTA) loss back-propagates through a soft-argmax of the coarse similarity map so every image patch receives gradient, not just sparsely sampled positive pairs. These losses sit inside a Fusion Transformer and are trained jointly with an in-frustum detection loss, coarse and fine matching losses, and finally EPnP-RANSAC pose estimation.

What would settle it

Evaluate TrafficLoc on a held-out real intersection where camera poses are surveyed with a total station, GNSS/IMU, or known calibration targets, after training only on Carla Intersection. If median rotation error on that real set is much worse than the 0.66 degrees seen in simulation, for example several degrees or more, the claimed sim-to-real generalization would be falsified even though the simulated numbers stand.

Watch

Extended reading notes

Core claim

The paper's central claim is that TrafficLoc, a neural network for image-to-point-cloud registration, localizes traffic cameras in 3D scenes with substantially higher accuracy than prior methods, and does so without requiring an initial pose guess, manual 2D-3D correspondences, or panoramic or rendered images. The method first extracts 2D patch features and 3D point-group features, fuses them with a Transformer whose cross-attention maps are supervised by a geometry-guided attention loss derived from camera projection, matches groups to patches in coarse-to-fine stages, and finally feeds point-to-pixel correspondences to EPnP-RANSAC to obtain the 6-DoF pose. On the paper's Carla Intersection test splits, median rotation error falls to 0.66 degrees and median translation to 0.51 m, versus 4.24 degrees and 2.82 m for the previous leading baseline CoFiI2P, an improvement the paper reports as up to 86% across test splits. On KITTI the mean rotation error is 0.87 degrees with 100% registration recall, and the paper also reports that the same model, trained only on simulation, produces visually correct alignment on a real traffic intersection with unknown intrinsics, and that adding its loss components to an existing registration method improves that method as well.

Load-bearing premise

The load-bearing premise is that the simulated Carla Intersection images are close enough to real traffic-camera views that a model trained only on them will localize real cameras; the paper's real-world demonstration is qualitative and has no ground-truth pose, so this assumption is not yet measured.

Editorial extensions

If this is right

  • Traffic cameras with known or estimated intrinsic parameters can be registered to an existing LiDAR map in a single forward pass, with no manual keypoint annotation and no panoramic capture.
  • Because the method is trained entirely on simulated intersections yet reports low errors on in-vehicle benchmarks, the same model may serve for both roadside and vehicle-mounted cameras, reducing dataset collection cost.
  • Sub-degree rotation and sub-metre translation on the simulated test split would be sufficient for many cooperative-perception tasks such as projecting detections from multiple cameras into a common coordinate frame.
  • The proposed losses are transferable: the paper shows that adding DTA and GAL to the previous leading method CoFiI2P improves its KITTI results, implying the components are not tied to TrafficLoc's specific architecture.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: If sim-to-real transfer holds, traffic-camera calibration becomes a city-scale mapping service rather than a per-camera manual procedure; cameras that already exist could be localized against maps captured by survey vehicles.
  • Editorial inference: The GAL formulation decouples attention supervision from appearance, so it could be applied to other cross-modal alignment tasks, such as radar-to-image or satellite-to-street matching, where geometry is known but photometry differs.
  • Editorial inference: The paper's public release of Carla Intersection with systematic pose sampling would let the community test viewpoint robustness of any image-to-point-cloud method independently of real-world sensor noise.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper addresses the problem of localizing traffic surveillance cameras within a 3D point cloud map by proposing TrafficLoc, a coarse-to-fine image-to-point-cloud (I2P) registration method. The method consists of a Geometry-guided Feature Fusion (GFF) module supervised by a Geometry-guided Attention Loss (GAL), an Inter-intra Contrastive Learning (ICL) loss for coarse matching, and a Dense Training Alignment (DTA) step with soft-argmax for fine matching. The authors also introduce Carla Intersection, a synthetic dataset of 75 intersections in the Carla simulator, and evaluate TrafficLoc on this dataset, on the real USTC intersection, and on the KITTI and NuScenes benchmarks. The central claims are that TrafficLoc greatly improves performance over previous I2P methods on Carla Intersection (up to 86% RRE reduction), achieves new state-of-the-art results on KITTI and NuScenes, and generalizes to real-world traffic camera data.

Significance. If the reported numbers are reproducible, this is a substantial contribution to I2P registration and to the specific application of traffic camera localization. The paper introduces a new synthetic benchmark that is likely to be useful to the community, and the proposed components (GAL, ICL, DTA) are technically plausible and are backed by extensive ablations in the main text and supplementary material. The authors state that code and data will be released and provide a public project page, which strengthens the reproducibility case. However, the evaluation protocol has several gaps that affect the strength of the central numerical claims: hyperparameters are selected on the test split, no error bars are reported, and the baseline comparisons are not fully controlled. These issues are fixable with additional experiments, and the underlying method appears sound, so the paper merits revision rather than rejection.

major comments (3)
  1. [Section 5.4 and Supplementary Table 4] The GAL hyperparameters (θlow, θup, dlow, dup, and the fusion layer to which GAL is applied) are selected on the TestT1−T7 split, which is also the split used for the main evaluation. Section 5.4 explicitly states that ablations are run on TestT1−T7, and Supplementary Table 4 reports threshold choices on that same split. This constitutes test-set tuning of the proposed components, while the baselines in Table 1 are not given the same opportunity; consequently, the reported margins over CoFiI2P (e.g., RRE 0.66° vs 4.24° on TestT1−T7) may partly reflect favorable hyperparameter selection. Please rerun the ablations on a held-out validation split (for example, a subset of the training intersections) and report the corresponding results, and provide variance estimates over multiple training runs for the final configuration.
  2. [Section 5.3, Table 1] The claim that TrafficLoc 'achieves new SOTA performance on KITTI and NuScenes' is supported by a single training run and by comparison to baselines that use different backbones and training configurations. No error bars are reported, and TrafficLoc uses a DUSt3R image encoder, a specific point-cloud size, and a four-layer fusion transformer, which differ from the original baseline setups. This makes it difficult to attribute the improvement to the proposed GAL/ICL/DTA components rather than to the stronger backbone or training-supervision differences. Please report mean and standard deviation over at least three seeds for TrafficLoc and the strongest baselines on KITTI and NuScenes, and provide a controlled comparison in which the baseline is given the same backbone and training procedure; Supplementary Table 5 partially addresses this for CoFiI2P, but the headline comparison in Table 1 remains uncontrolled.
  3. [Section 5.3, Figure 5] The real-world generalization claim is only qualitatively demonstrated on the USTC intersection, where no ground-truth pose is available. Since the abstract states that TrafficLoc 'generalizes well to real-world data', the absence of any quantitative real-world evaluation leaves this part of the central claim unsupported. Please provide a quantitative evaluation (for example, against manually annotated or survey-grade poses) or, if that is not feasible, explicitly weaken the claim in the abstract and introduction to a qualitative demonstration.
minor comments (5)
  1. [Section 4.1] There is a typo: 'describied' should be 'described'.
  2. [Section 4.2, Eqs. (2)-(3)] The notation OIi and OPj is not defined in the text; please clarify that Ii and Pj denote the 3D positions of the image patch and point-group center relative to the camera center O, or add a small figure illustrating the angular radius Rad(i,j) and the distance Dist(Pi, OIj).
  3. [Table 1] The abbreviation 'Ours withKP red' is unclear; please replace it with 'Ours with predicted K' or 'Ours with Kpred'.
  4. [References] Several reference entries contain stray page numbers or suffixes (for example, references [5], [13], and [14]), and the capitalization of 'nuScenes' is inconsistent; please clean up the bibliography.
  5. [Throughout] The spelling 'Nuscenes' is used inconsistently; the official spelling is 'nuScenes'.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: TrafficLoc's reported improvements come from supervised training and held-out evaluation, not from fitted parameters or self-citation chains.

full rationale

The derivation chain is self-contained: TrafficLoc is trained with standard supervised losses (GAL, ICL, DTA) that use ground-truth poses and correspondences only during training, while the reported Table 1 numbers are measured on held-out Carla Intersection test splits (unseen intersections and, for TestT10, an unseen world), on KITTI sequences 9-10, and on the NuScenes test split. At inference the pose is produced by EPnP-RANSAC from predicted point-to-pixel correspondences, so the accuracy does not reduce to any fitted parameter by construction. The only in-article self-citations ([26], [49], [50]) appear as background references for place recognition and LiDAR utility; none is load-bearing for the design or the claimed SOTA, and no uniqueness theorem is imported from the authors' prior work. The GAL supervision target is derived from camera projection geometry plus known intrinsics, which is a training signal, not a test-time prediction. The USTC qualitative projection is a sanity check rather than the basis of a numeric derivation; it does not support a quantitative SOTA claim on its own, but it is also not a circular reduction. Supplementary Table 5 further shows that DTA and GAL improve the external CoFiI2P baseline on KITTI, which is independent evidence that these components are not merely fitting TrafficLoc's own pipeline. Concerns about threshold selection on the test split (Table 4) and missing variance estimates are evaluation-protocol weaknesses, not circular reductions; with no equation or fitted value renamed as a prediction, no circular step is exhibited.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central claim rests on a supervised training pipeline that assumes access to ground-truth poses and known camera geometry during training, plus a set of hyperparameters (GAL thresholds, contrastive margins, loss weights) that are chosen by hand or via ablation. The method does not introduce new physical entities. The main unstated risk is the representativeness of the synthetic dataset, captured in the weakest_assumption field.

free parameters (4)
  • GAL thresholds (θlow, θup, dlow, dup) = 10°, 20°, 3 m, 5 m
    Selected via ablation on the Carla test split (Supplementary Table 4); different values change RRE/RTE noticeably, so reported performance is partly tuned to the test data.
  • ICL margins and scale (mp, mn, γ) = 0.2, 1.8, 10
    Set by hand; control the separation of positive and negative pairs in the Inter-intra Contrastive Loss.
  • Loss weights (λ1, λ2, λ3, λ4) = 1, 1, 1, 1
    Set to equal contribution without tuning.
  • Input point cloud size (N) = 20480
    Ablated in Supplementary Table 9; 20480 balances accuracy and FLOPs.
assumptions (4)
  • domain assumption Pinhole camera model with known intrinsic matrix K (or K estimated by DUSt3R)
    Invoked in Section 3 and used throughout for projecting 3D points into the image; the DUSt3R estimation is itself a learned prior.
  • domain assumption Ground-truth 6-DoF pose TGT is available for every training image
    Used to define GAL attention targets, ICL positive/negative pairs, and DTA supervision (Sections 4.2, 4.3, 4.6). This is standard supervised training but is a requirement.
  • domain assumption The image and point cloud overlap sufficiently for matching
    Supplementary Section E pairs a voxel with an image only if more than 30% of projected points fall inside the image plane, so the method assumes meaningful overlap.
  • standard math EPnP-RANSAC recovers the correct pose from a set of point-to-pixel correspondences
    Standard geometric solver used in prior I2P work [20, 24]; the paper relies on it for the final pose.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TrafficLoc: Localizing Traffic Surveillance Cameras in 3D Scenes." pith.science (2026). https://pith.science/paper/R36IUKOY

@misc{pith2026241210308,
  author       = {Pith},
  title        = {Pith review of: TrafficLoc: Localizing Traffic Surveillance Cameras in 3D Scenes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/R36IUKOY}},
  note         = {Machine review of arXiv:2412.10308}
}
read the original abstract

We tackle the problem of localizing traffic cameras within a 3D reference map and propose a novel image-to-point cloud registration (I2P) method, TrafficLoc, in a coarse-tofine matching fashion. To overcome the lack of large-scale real-world intersection datasets, we first introduce Carla Intersection, a new simulated dataset with 75 urban and rural intersections in Carla. We find that current I2P methods struggle with cross-modal matching under large viewpoint differences, especially at traffic intersections. TrafficLoc thus employs a novel Geometry-guided Attention Loss (GAL) to focus only on the corresponding geometric regions under different viewpoints during 2D-3D feature fusion. To address feature inconsistency in paired image patch-point groups, we further propose Inter-intra Contrastive Learning (ICL) to enhance separating 2D patch/3D group features within each intra-modality and introduce Dense Training Alignment (DTA) with soft-argmax for improving position regression. Extensive experiments show our TrafficLoc greatly improves the performance over the SOTA I2P methods (up to 86%) on Carla Intersection and generalizes well to real-world data. TrafficLoc also achieves new SOTA performance on KITTI and NuScenes datasets, demonstrating the superiority across both in-vehicle and traffic cameras. Our project page is publicly available at https://tum-luk.github.io/projects/trafficloc/.

Figures

Figures reproduced from arXiv: 2412.10308 by the authors.

Figure 1
Figure 1. Localization accuracy on the proposed Carla Intersection and KITTI dataset. The point cloud is projected into a 2D view and shown above the image, with point colors indicating distance. The proposed TrafficLoc achieves better performance, with more correct (green) and fewer incorrect (red) point-to-pixel pairs. The first column presents the input point cloud and input image. Abstract We tackle the problem of localiz… view at source ↗
Figure 2
Figure 2. Pipeline of our TrafficLoc. Given a traffic camera image and a 3D scene point cloud collected at different locations, we first [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The pipeline of Geometry-guided Feature Fusion (GFF) module. GFF first use [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Coarse matching mechanism of TrafficLoc. The positive [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Localization performance of our TrafficLoc on the [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Visualization result of using different loss function. (a), [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Visualization result of P2I and I2P attention map when [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Sampled testing image poses of (a) Town01 Intersec [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]
Figure 9
Figure 9. Figure 9: Point cloud capturing example from T own10 Int1. (a1) and (a2) depict the LiDAR scan from a single frame. (b) shows the aggregated and downsampled point cloud. (c) presents the final cropped point cloud with dimensions of 100 m×100 m×100 m. GT Projection 𝟖. 𝟔𝟒°/𝟐. 𝟑𝟖𝒎 …
Figure 10
Figure 10. Figure 10: Qualitative results of our TrafficLoc and other baseline methods on the KITTI Odometry dataset [ [PITH_FULL_IMAGE:figures/full_fig_p016_10.png]
Figure 11
Figure 11. Figure 11: Example point clouds and images data of our [PITH_FULL_IMAGE:figures/full_fig_p018_11.png]
Figure 12
Figure 12. Figure 12: Qualitative results of our TrafficLoc and other baseline methods on the [PITH_FULL_IMAGE:figures/full_fig_p019_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 51 canonical work pages

  1. [1]

    Building rome in a day

    Sameer Agarwal, Yasutaka Furukawa, Noah Snavely, Ian Si- mon, Brian Curless, Steven M Seitz, and Richard Szeliski. Building rome in a day. Communications of the ACM , 54 (10):105–112, 2011. 2

    2011

  2. [2]

    Buffer: Balancing accuracy, efficiency, and gen- eralizability in point cloud registration

    Sheng Ao, Qingyong Hu, Hanyun Wang, Kai Xu, and Yu- lan Guo. Buffer: Balancing accuracy, efficiency, and gen- eralizability in point cloud registration. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 1255–1264, 2023. 3

    work page 2023

  3. [3]

    Method for registration of 3-d shapes

    Paul J Besl and Neil D McKay. Method for registration of 3-d shapes. In Sensor fusion IV: control paradigms and data structures, pages 586–606. International Society for Optics and Photonics, 1992. 2

    work page 1992

  4. [4]

    A light touch approach to teaching transformers multi-view geometry

    Yash Bhalgat, Joao F Henriques, and Andrew Zisserman. A light touch approach to teaching transformers multi-view geometry. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 4958– 4969, 2023. 4

    work page 2023

  5. [5]

    Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Gi- ancarlo Baldan, and Oscar Beijbom

    Holger Caesar, Varun Bankiti, Alex H. Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Gi- ancarlo Baldan, and Oscar Beijbom. nuscenes: A multi- modal dataset for autonomous driving. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020. 2, 7, 5, 6

    work page 2020

  6. [6]

    Global visual localization in lidar-maps through shared 2d-3d embedding space

    Daniele Cattaneo, Matteo Vaghi, Simone Fontana, Au- gusto Luis Ballardini, and Domenico Giorgio Sorrenti. Global visual localization in lidar-maps through shared 2d-3d embedding space. In International Conference on Robotics and Automation (ICRA) , pages 4365–4371, 2020. 3

    work page 2020

  7. [7]

    Pimae: Point cloud and image interactive masked autoencoders for 3d object detection

    Anthony Chen, Kevin Zhang, Renrui Zhang, Zihan Wang, Yuheng Lu, Yandong Guo, and Shanghang Zhang. Pimae: Point cloud and image interactive masked autoencoders for 3d object detection. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 5291–5301, 2023. 1, 2

    work page 2023

  8. [8]

    Object modeling by regis- tration of multiple range images

    Yang Chen and G ´erard Medioni. Object modeling by regis- tration of multiple range images. Image and Vision Comput- ing, 10(3):145–155, 1992. 3

    work page 1992

Show all 60 references
  1. [9]

    Superpoint: Self-supervised interest point detec- tion and description

    Daniel DeTone, Tomasz Malisiewicz, and Andrew Rabi- novich. Superpoint: Self-supervised interest point detec- tion and description. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition Work- shops (CVPRW), pages 224–236, 2018. 2

    2018

  2. [10]

    Carla: An open urban driv- ing simulator

    Alexey Dosovitskiy, German Ros, Felipe Codevilla, Anto- nio L´opez, and Vladlen Koltun. Carla: An open urban driv- ing simulator. Conference on Robot Learning,Conference on Robot Learning, 2017. 6, 3

    2017

  3. [11]

    D2-net: A trainable cnn for joint detection and description of local features

    Mihai Dusmanu, Ignacio Rocco, Tomas Pajdla, Marc Polle- feys, Josef Sivic, Akihiko Torii, and Torsten Sattler. D2-net: A trainable cnn for joint detection and description of local features. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (C...

    2019

  4. [12]

    2d3d-matchnet: Learning to match keypoints across 2d image and 3d point cloud

    Mengdan Feng, Sixing Hu, Marcelo H Ang, and Gim Hee Lee. 2d3d-matchnet: Learning to match keypoints across 2d image and 3d point cloud. In International Conference on Robotics and Automation (ICRA) , pages 4790–4796. IEEE,

  5. [13]

    Fischler and Robert C

    Martin A. Fischler and Robert C. Bolles. Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography. Commun. ACM, 24(6):381–395, 1981. 3, 4, 6, 7, 2, 5

    1981

  6. [14]

    Are we ready for autonomous driving? the kitti vision benchmark suite

    Andreas Geiger, Philip Lenz, and Raquel Urtasun. Are we ready for autonomous driving? the kitti vision benchmark suite. In Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), 2012. 2, 7, 1, 5, 6

    2012

  7. [15]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceed- ings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016. 3, 6, 1, 2

    2016

  8. [16]

    Anpr product functions and problem troubleshooting

    HIKVISION. Anpr product functions and problem troubleshooting. https : / / www . securitywholesalers . com . au / files / ANPRINSTALLATION1.pdf, 2016–2024. 3

    2016

  9. [17]

    Predator: Registration of 3d point clouds with low overlap

    Shengyu Huang, Zan Gojcic, Mikhail Usvyatsov, Andreas Wieser, and Konrad Schindler. Predator: Registration of 3d point clouds with low overlap. In Proceedings of the IEEE/CVF Conference on computer vision and pattern recognition, pages 4267–4276, 2021. 3

    2021

  10. [18]

    Reconstructing the world in six days

    Heinly Jared, Johannes L Schonberger, Enrique Dunn, and Jan-Michael Frahm. Reconstructing the world in six days. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2015. 2

    2015

  11. [19]

    Intrinsic and extrinsic calibration of roadside lidar and camera

    Xin Jing, Fuzhang Han, Xiaqing Ding, Yue Wang, and Rong Xiong. Intrinsic and extrinsic calibration of roadside lidar and camera. In 2022 China Automation Congress (CAC) , pages 2367–2372. IEEE, 2022. 2

    2022

  12. [20]

    Cofii2p: Coarse-to-fine correspon- dences for image-to-point cloud registration

    Shuhao Kang, Youqi Liao, Jianping Li, Fuxun Liang, Yuhao Li, Xianghong Zou, Fangning Li, Xieyuanli Chen, Zhen Dong, and Bisheng Yang. Cofii2p: Coarse-to-fine correspon- dences for image-to-point cloud registration. arXiv preprint arXiv:2309.14660, 2023. 2, 3, 5, 6, 7, 8, 1

    2023 arXiv

  13. [21]

    Ep2p-loc: End-to-end 3d point to 2d pixel localization for large-scale visual localization

    Minjung Kim, Junseo Koo, and Gunhee Kim. Ep2p-loc: End-to-end 3d point to 2d pixel localization for large-scale visual localization. In Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision, pages 21527–21537,

  14. [22]

    Kingma and Jimmy Lei Ba

    Diederik P. Kingma and Jimmy Lei Ba. Adam: A method for stochastic optimization. In International Conference on Learning Representations, 2015. 5

    2015

  15. [23]

    General, single-shot, target-less, and automatic lidar-camera extrinsic calibration toolbox

    Kenji Koide, Shuji Oishi, Masashi Yokozuka, and Atsuhiko Banno. General, single-shot, target-less, and automatic lidar-camera extrinsic calibration toolbox. arXiv preprint arXiv:2302.05094, 2023. 2

    2023 arXiv

  16. [24]

    Epnp: An accurate o(n) solution to the pnp problem

    Vincent Lepetit, Francesc Moreno-Noguer, and Pascal Fua. Epnp: An accurate o(n) solution to the pnp problem. Inter- national Journal of Computer Vision (IJCV), 81(2):155–166,

  17. [25]

    DeepI2P: Image-to-Point Cloud Registration via Deep Classification

    Jiaxin Li and Gim Hee Lee. DeepI2P: Image-to-Point Cloud Registration via Deep Classification. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 15960–15969, 2021. 3, 7, 5 9

    2021

  18. [26]

    Vxp: V oxel-cross-pixel large-scale image- lidar place recognition

    Yun-Jin Li, Mariia Gladkova, Yan Xia, Rui Wang, and Daniel Cremers. Vxp: V oxel-cross-pixel large-scale image- lidar place recognition. In 2025 International Conference on 3D Vision (3DV), 2025. 3

    2025

  19. [27]

    Lightglue: Local feature matching at light speed

    Philipp Lindenberger, Paul-Edouard Sarlin, and Marc Polle- feys. Lightglue: Local feature matching at light speed. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 17627–17638, 2023. 2

    2023

  20. [28]

    Extend your own correspondences: Unsupervised distant point cloud registration by progressive distance extension

    Quan Liu, Hongzi Zhu, Zhenxi Wang, Yunsong Zhou, Shan Chang, and Minyi Guo. Extend your own correspondences: Unsupervised distant point cloud registration by progressive distance extension. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition,...

    2024

  21. [29]

    Distinctive image features from scale- invariant keypoints

    David G Lowe. Distinctive image features from scale- invariant keypoints. International Journal of Computer Vi- sion (IJCV), 60(2):91–110, 2004. 2

    2004

  22. [30]

    Deepicp: An end- to-end deep neural network for 3d point cloud registration

    W Lu, G Wan, Y Zhou, X Fu, and S Song. Deepicp: An end- to-end deep neural network for 3d point cloud registration. arXiv preprint arXiv:1905.04153, 2019. 3

    1905 arXiv

  23. [31]

    Orb-slam: A versatile and accurate monocular slam system

    Raul Mur-Artal, Jose Maria Martinez Montiel, and Juan D Tardos. Orb-slam: A versatile and accurate monocular slam system. IEEE transactions on robotics , 31(5):1147–1163,

  24. [32]

    A survey of structure from motion

    Onur Ozyesil, Vladislav V oroninski, Ronen Basri, and Amit Singer. A survey of structure from motion. arXiv preprint arXiv:1701.08493, 2017. 2

    2017 arXiv

  25. [33]

    Lcd: Learned cross-domain descriptors for 2d-3d matching

    Quang-Hieu Pham, Mikaela Angelina Uy, Binh-Son Hua, Duc Thanh Nguyen, Gemma Roig, and Sai-Kit Yeung. Lcd: Learned cross-domain descriptors for 2d-3d matching. In Proceedings of the AAAI Conference on Artificial Intelli- gence (AAAI), pages 11856–11864, 2020. 3

    2020

  26. [34]

    Pointnet: Deep learning on point sets for 3d classification and segmentation

    Charles R Qi, Hao Su, Kaichun Mo, and Leonidas J Guibas. Pointnet: Deep learning on point sets for 3d classification and segmentation. In Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition (CVPR) , pages 652–660, 2017. 3, 6

    2017

  27. [35]

    Geotrans- former: Fast and robust point cloud registration with geo- metric transformer

    Zheng Qin, Hao Yu, Changjian Wang, Yulan Guo, Yuxing Peng, Slobodan Ilic, Dewen Hu, and Kai Xu. Geotrans- former: Fast and robust point cloud registration with geo- metric transformer. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(8):9806–9821, 2023. 3

    2023

  28. [36]

    Real-time dy- namic object detection for autonomous driving using prior 3d-maps

    B Ravi Kiran, Luis Roldao, Benat Irastorza, Renzo Ve- rastegui, Sebastian Suss, Senthil Yogamani, Victor Talpaert, Alexandre Lepoutre, and Guillaume Trehard. Real-time dy- namic object detection for autonomous driving using prior 3d-maps. In Proceedings of the European Confere...

    2018

  29. [37]

    Corri2p: Deep image-to-point cloud registration via dense correspondence

    Siyu Ren, Yiming Zeng, Junhui Hou, and Xiaodong Chen. Corri2p: Deep image-to-point cloud registration via dense correspondence. IEEE Transactions on Circuits and Systems for Video Technology, 33(3):1198–1208, 2022. 2, 3, 7, 5

    2022

  30. [38]

    R2D2: repeatable and re- liable detector and descriptor

    Jerome Revaud, Philippe Weinzaepfel, C ´esar Roberto de Souza, and Martin Humenberger. R2D2: repeatable and re- liable detector and descriptor. In Advances in Neural Infor- mation Processing Systems (NeurIPS), 2019. 2

    2019

  31. [39]

    Orb: An efficient alternative to sift or surf

    Ethan Rublee, Vincent Rabaud, Kurt Konolige, and Gary Bradski. Orb: An efficient alternative to sift or surf. In 2011 International conference on computer vision , pages 2564–

    2011

  32. [40]

    Efficient variants of the icp algorithm

    Szymon Rusinkiewicz and Marc Levoy. Efficient variants of the icp algorithm. In Proceedings third international confer- ence on 3-D digital imaging and modeling , pages 145–152. IEEE, 2001. 3

    2001

  33. [41]

    SuperGlue: Learning feature matching with graph neural networks

    Paul-Edouard Sarlin, Daniel DeTone, Tomasz Malisiewicz, and Andrew Rabinovich. SuperGlue: Learning feature matching with graph neural networks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020. 2

    2020

  34. [42]

    Rendering-enhanced auto- matic image-to-point cloud registration for roadside scenes

    Yu Sheng, Lu Zhang, Xingchen Li, Yifan Duan, Yanyong Zhang, Yu Zhang, and Jianmin Ji. Rendering-enhanced auto- matic image-to-point cloud registration for roadside scenes. arXiv preprint arXiv:2404.05164, 2024. 2, 7, 8

    2024 arXiv

  35. [43]

    Loftr: Detector-free local feature matching with transformers

    Jiaming Sun, Zehong Shen, Yuang Wang, Hujun Bao, and Xiaowei Zhou. Loftr: Detector-free local feature matching with transformers. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 8922–8931, 2021. 2

    2021

  36. [44]

    Neumap: Neural coordinate map- ping by auto-transdecoder for camera localization

    Shitao Tang, Sicong Tang, Andrea Tagliasacchi, Ping Tan, and Yasutaka Furukawa. Neumap: Neural coordinate map- ping by auto-transdecoder for camera localization. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 929–939, 2023. 4

    2023

  37. [45]

    Toward planet-wide traffic camera cali- bration

    Khiem Vuong, Robert Tamburo, and Srinivasa G Narasimhan. Toward planet-wide traffic camera cali- bration. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision , pages 8553–8562,

  38. [46]

    Freereg: Image- to-point cloud registration leveraging pretrained diffusion models and monocular depth estimators

    Haiping Wang, Yuan Liu, Bing Wang, Yujing Sun, Zhen Dong, Wenping Wang, and Bisheng Yang. Freereg: Image- to-point cloud registration leveraging pretrained diffusion models and monocular depth estimators. arXiv preprint arXiv:2310.03420, 2023. 3

    arXiv

  39. [47]

    Dust3r: Geometric 3d vi- sion made easy

    Shuzhe Wang, Vincent Leroy, Yohann Cabon, Boris Chidlovskii, and Jerome Revaud. Dust3r: Geometric 3d vi- sion made easy. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 20697– 20709, 2024. 3, 7, 1, 2, 4

    2024

  40. [48]

    Efficient loftr: Semi-dense local feature matching with sparse-like speed

    Yifan Wang, Xingyi He, Sida Peng, Dongli Tan, and Xiaowei Zhou. Efficient loftr: Semi-dense local feature matching with sparse-like speed. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 21666–21675, 2024. 2

    2024

  41. [49]

    Soe-net: A self-attention and orientation encoding network for point cloud based place recognition

    Yan Xia, Yusheng Xu, Shuang Li, Rui Wang, Juan Du, Daniel Cremers, and Uwe Stilla. Soe-net: A self-attention and orientation encoding network for point cloud based place recognition. In Proceedings of the IEEE/CVF Conference on computer vision and pattern recognition , pages 1...

    2021

  42. [50]

    Casspr: Cross attention single scan place recognition

    Yan Xia, Mariia Gladkova, Rui Wang, Qianyun Li, Uwe Stilla, Joao F Henriques, and Daniel Cremers. Casspr: Cross attention single scan place recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vi- sion, pages 8461–8472, 2023. 2

    2023

  43. [51]

    A lightweight and detector-free 3d single object 10 tracker on point clouds

    Yan Xia, Qiangqiang Wu, Wei Li, Antoni B Chan, and Uwe Stilla. A lightweight and detector-free 3d single object 10 tracker on point clouds. IEEE Transactions on Intelligent Transportation Systems, 24(5):5543–5554, 2023. 1

    2023

  44. [52]

    3dtnet: Learning local features us- ing 2d and 3d cues

    Xiaoxia Xing, Yinghao Cai, Tao Lu, Shaojun Cai, Yiping Yang, and Dayong Wen. 3dtnet: Learning local features us- ing 2d and 3d cues. InInternational Conference on 3D Vision (3DV), pages 435–443, 2018. 3

    2018

  45. [53]

    Cfi2p: Coarse-to-fine cross-modal correspondence learn- ing for image-to-point cloud registration

    Gongxin Yao, Yixin Xuan, Yiwei Chen, and Yu Pan. Cfi2p: Coarse-to-fine cross-modal correspondence learn- ing for image-to-point cloud registration. arXiv preprint arXiv:2307.07142, 2023. 2, 3, 4, 5, 7

    2023 arXiv

  46. [54]

    Cofinet: Reliable coarse-to-fine correspondences for robust pointcloud registration

    Hao Yu, Fu Li, Mahdi Saleh, Benjamin Busam, and Slobo- dan Ilic. Cofinet: Reliable coarse-to-fine correspondences for robust pointcloud registration. Advances in Neural Infor- mation Processing Systems, 34:23872–23884, 2021. 3

    2021

  47. [55]

    Pixel-level extrinsic self calibration of high resolution lidar and camera in targetless environments

    Chongjian Yuan, Xiyuan Liu, Xiaoping Hong, and Fu Zhang. Pixel-level extrinsic self calibration of high resolution lidar and camera in targetless environments. IEEE Robotics and Automation Letters, 6(4):7517–7524, 2021. 2

    2021

  48. [56]

    3dmatch: Learning local geometric descriptors from rgb-d reconstruc- tions

    Andy Zeng, Shuran Song, Matthias Niessner, Matthew Fisher, Jianxiong Xiao, and Thomas Funkhouser. 3dmatch: Learning local geometric descriptors from rgb-d reconstruc- tions. In Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), 2017. 3

    2017

  49. [57]

    Telling left from right: Identifying geometry-aware semantic corre- spondence

    Junyi Zhang, Charles Herrmann, Junhwa Hur, Eric Chen, Varun Jampani, Deqing Sun, and Ming-Hsuan Yang. Telling left from right: Identifying geometry-aware semantic corre- spondence. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 307...

    2024

  50. [58]

    Point transformer

    Hengshuang Zhao, Li Jiang, Jiaya Jia, Philip HS Torr, and Vladlen Koltun. Point transformer. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 16259–16268, 2021. 3, 1, 2

    2021

  51. [59]

    NCL” means using normal contrastive learning, while “ICL

    Junsheng Zhou, Baorui Ma, Wenyuan Zhang, Yi Fang, Yu- Shen Liu, and Zhizhong Han. Differentiable registration of images and lidar point clouds with voxelpoint-to-pixel matching. Advances in Neural Information Processing Sys- tems, 36, 2024. 3, 7, 8 11 TrafficLoc: Localizing Tr...

    2024

  52. [960]

    # intersections

    To generate these images, we sampled camera posi- tions in a grid-like pattern with different heights at the cen- ter of each intersection. For each position, we captured im- ages at 8 yaw angles (spaced at 45 ◦ intervals) and 2 pitch angles. Figure 8 shows the sampled poses f...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.