Exploiting uncertainty in regression forests for accurate camera relocalization
The DOI matches its Crossref record; 23 of the 23 references it cites were found in a registry or an index.
Abstract
Recent advances in camera relocalization use predictions from a regression forest to guide the camera pose optimization procedure. In these methods, each tree associates one pixel with a point in the scene's 3D world coordinate frame. In previous work, these predictions were point estimates and the subsequent camera pose optimization implicitly assumed an isotropic distribution of these estimates. In this paper, we train a regression forest to predict mixtures of anisotropic 3D Gaussians and show how the predicted uncertainties can be taken into account for continuous pose optimization. Experiments show that our proposed method is able to relocalize up to 40% more frames than the state of the art.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Likewise it is possible to localize the 6DoF of a camera with regression forest [247]”
From Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age (2016) · cites this paper as [247]
Citations
Cites
Mean shift: a robust approach toward feature space analysis
Comaniciu, D., & Meer, P. (2002) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/34.1000236 · Cited by 11,305 · cited 1×
“Similar to [19], we run mean shift mode detection [3] to find the modes of the empirical distribution in each leaf.”
MethodKinectFusion: Real-time dense surface mapping and tracking
Newcombe, R. A., Fitzgibbon, A., Izadi, S., Hilliges, O., Molyneaux, D., Kim, D., Davison, A. J., Kohi, P., Shotton, J., & Hodges, S. (2011) · 2011 10th IEEE International Symposium on Mixed and Augmented Reality · doi:10.1109/ismar.2011.6092378 · Cited by 4,008 · cited 1×
“Simultaneous Localization and Mapping (SLAM) systems such as [14] [15] [17] [7] and commercial systems such as HoloLens or Project Tango demonstrate that visual SLAM is a maturing technology.”
BackgroundSelf-citationLSD-SLAM: Large-scale direct monocular SLAM
Engel, J., Schöps, T., & Cremers, D. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10605-2_54 · Cited by 3,851 · cited 1×
“Simultaneous Localization and Mapping (SLAM) systems such as [14] [15] [17] [7] and commercial systems such as HoloLens or Project Tango demonstrate that visual SLAM is a maturing technology.”
BackgroundA solution for the best rotation to relate two sets of vectors
Kabsch, W. (1976) · Acta Crystallographica Section A · doi:10.1107/s0567739476001873 · Cited by 3,187 · cited 1×
“When per-pixel depth information is available, one can estimate the transformation parameters from the current view to the model with three local positions (current camera space) and their corresponding positions in the scene (world coordinates) using the Kabsch algorithm [11].”
MethodDTAM: Dense tracking and mapping in real-time
Newcombe, R. A., Lovegrove, S. J., & Davison, A. J. (2011) · 2011 International Conference on Computer Vision · doi:10.1109/iccv.2011.6126513 · Cited by 2,078 · cited 1×
“Simultaneous Localization and Mapping (SLAM) systems such as [14] [15] [17] [7] and commercial systems such as HoloLens or Project Tango demonstrate that visual SLAM is a maturing technology.”
BackgroundReal-time human pose recognition in parts from single depth images
Shotton, J., Sharp, T., Kipman, A., Fitzgibbon, A., Finocchio, M., Blake, A., Cook, M., & Moore, R. (2013) · Communications of the ACM · doi:10.1145/2398356.2398381 · Cited by 2,044 · cited 1×
“Following [19], we use features based on pairwise pixel relationships [13, 20] as they have proven to be well-suited for the task of camera relocalization.”
MethodSelf-citationFAB-MAP: Probabilistic Localization and Mapping in the Space of Appearance
Cummins, M., & Newman, P. (2008) · The International Journal of Robotics Research · doi:10.1177/0278364908090961 · Cited by 1,472 · cited 1×
“A few successful approaches also exist that do not belong to either of those two classes [5, 16].”
BackgroundReal-time 3D reconstruction at scale using voxel hashing
Nießner, M., Zollhöfer, M., Izadi, S., & Stamminger, M. (2013) · ACM Transactions on Graphics · doi:10.1145/2508363.2508374 · Cited by 1,002 · cited 1×
“Simultaneous Localization and Mapping (SLAM) systems such as [14] [15] [17] [7] and commercial systems such as HoloLens or Project Tango demonstrate that visual SLAM is a maturing technology.”
BackgroundSelf-citationScene Coordinate Regression Forests for Camera Relocalization in RGB-D Images
Shotton, J., Glocker, B., Zach, C., Izadi, S., Criminisi, A., & Fitzgibbon, A. (2013) · 2013 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2013.377 · Cited by 930 · cited 3×
“Perhaps the simplest approach, used in [19], is to fit an isotropic Gaussian to the labels in S.”
BackgroundMethodInfluentialSelf-citationChum, O., Matas, J., & Kittler, J. (2003) · Lecture Notes in Computer Science · doi:10.1007/978-3-540-45243-0_31 · Cited by 843 · cited 3×
“Having computed the set of inliers, [2] suggests performing a local optimization step to refine the hypotheses.”
MethodInfluential
Cited by
Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age
Cadena, C., Carlone, L., Carrillo, H., Latif, Y., Scaramuzza, D., Neira, J., Reid, I., & Leonard, J. J. (2016) · IEEE Transactions on Robotics · doi:10.1109/tro.2016.2624754 · Cited by 3,596
“Likewise, it is possible to localize the 6DoF of a camera with regression forest [247] and with deep convolutional neural network [129], and to estimate the depth of a scene (in effect, the map) from a single view solely as a function of the”
Past, present, and future of simultaneous localization and mapping: Toward the robust-perception age
Cadena, C., Carlone, L., Carrillo, H., Latif, Y., Scaramuzza, D., Neira, J., Reid, I. D., & Leonard, J. J. (2016) · Zurich Open Repository and Archive (University of Zurich) · doi:10.5167/uzh-138892 · Cited by 3,048
Normalized Object Coordinate Space for Category-Level 6D Object Pose and Size Estimation
Wang, H., Sridhar, S., Huang, J., Valentin, J., Song, S., & Guibas, L. J. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00275 · Cited by 757
Self-citationBundleFusion: Real-Time Globally Consistent 3D Reconstruction Using On-the-Fly Surface Reintegration
Dai, A., Nießner, M., Zollhöfer, M., Izadi, S., & Theobalt, C. (2017) · ACM Transactions on Graphics · doi:10.1145/3072959.3054739 · Cited by 662
Self-citationDSAC — Differentiable RANSAC for Camera Localization
Brachmann, E., Krull, A., Nowozin, S., Shotton, J., Michel, F., Gumhold, S., & Rother, C. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.267 · Cited by 607
“We demonstrate an end-to-end trainable solution for this problem, building on the scene coordinate regression forest (SCoRF) approach [36, 40, 5].”
MethodSelf-citationModelling uncertainty in deep learning for camera relocalization
Kendall, A., & Cipolla, R. (2016) · 2016 IEEE International Conference on Robotics and Automation (ICRA) · doi:10.1109/icra.2016.7487679 · Cited by 548
“It was extended with a probabilistic approach in [17].”
Uncertainty-Driven 6D Pose Estimation of Objects and Scenes from a Single RGB Image
Brachmann, E., Michel, F., Krull, A., Yang, M. Y., Gumhold, S., & Rother, C. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.366 · Cited by 530
“1% of correct poses, which is on-par with the results of the state-of-theart RGB-D method [29] (89.”
BackgroundMethodResultInfluentialImage-Based Localization Using LSTMs for Structured Feature Correlation
Walch, F., Hazirbas, C., Leal-Taixe, L., Sattler, T., Hilsenbeck, S., & Cremers, D. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.75 · Cited by 521
“explicitly model the uncertainty of the predicted 3D point positions and use this uncertainty during pose estimation [52], allowing them to localize more images .”
BackgroundiNeRF: Inverting Neural Radiance Fields for Pose Estimation
Yen-Chen, L., Florence, P., Barron, J. T., Rodriguez, A., Isola, P., & Lin, T.-Y. (2021) · 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros51168.2021.9636708 · Cited by 404
“[33], [41], [31]), because NeRF and iNeRF only require posed RGB images as training, iNeRF can be applied to localization as well.”
Learning Less is More - 6D Camera Localization via 3D Surface Regression
Brachmann, E., & Rother, C. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00489 · Cited by 399
State of the Art on 3D Reconstruction with RGB‐D Cameras
Zollhöfer, M., Stotko, P., Görlitz, A., Theobalt, C., Nießner, M., Klein, R., & Kolb, A. (2018) · Computer Graphics Forum · doi:10.1111/cgf.13386 · Cited by 339
Schonberger, J. L., Pollefeys, M., Geiger, A., & Sattler, T. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00721 · Cited by 267
Neural-Guided RANSAC: Learning Where to Sample Model Hypotheses
Brachmann, E., & Rother, C. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00442 · Cited by 260
“Notable exceptions include geometric problems like object instance pose estimation [3, 5, 21], and camera re-localization [41, 51, 28, 8, 46] where RANSAC is coupled with decision forests or neural networks that predict image-to-object correspondences.”
BackgroundBundleFusion: Real-Time Globally Consistent 3D Reconstruction Using On-the-Fly Surface Reintegration
Dai, A., Nießner, M., Zollhöfer, M., Izadi, S., & Theobalt, C. (2017) · ACM Transactions on Graphics · doi:10.1145/3054739 · Cited by 258
Self-citationDeep Auxiliary Learning for Visual Localization and Odometry
Valada, A., Radwan, N., & Burgard, W. (2018) · 2018 IEEE International Conference on Robotics and Automation (ICRA) · doi:10.1109/icra.2018.8462979 · Cited by 241
A survey on Visual-Based Localization: On the benefit of heterogeneous data
Piasco, N., Sidibé, D., Demonceaux, C., & Gouet-Brunet, V. (2018) · Pattern Recognition · doi:10.1016/j.patcog.2017.09.013 · Cited by 236
“Compared to regression forest [204], CNN seems more appropriate to handle large environment and does not rely on depth information.”
BackgroundRelocNet: Continuous Metric Learning Relocalisation Using Neural Nets
Balntas, V., Li, S., & Prisacariu, V. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01264-9_46 · Cited by 223
“Random forest based methods [42, 17, 45] deliver state-of-the-art accuracy, by regressing the camera location for each point in an RGBD query frame.”
MethodVisual Camera Re-Localization from RGB and RGB-D Images Using DSAC
Brachmann, E., & Rother, C. (2021) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2021.3070754 · Cited by 217
2D3D-Matchnet: Learning To Match Keypoints Across 2D Image And 3D Point Cloud
Feng, M., Hu, S., Ang, M. H., & Lee, G. H. (2019) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2019.8794415 · Cited by 213
“[16], [17] learn to predict pixel-wise image to scene correspondences, followed by a RANSAC-like optimization algorithm for pose estimation.”
MethodCamNet: Coarse-to-Fine Retrieval for Camera Re-Localization
Ding, M., Wang, Z., Sun, J., Shi, J., & Luo, P. (2019) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00296 · Cited by 212
“Some methods based on Random Forest [9, 23, 35, 40] establish 2D-3D matches to re- cover 6-DoF camera pose by applying RANSAC to achieve impressing results.”
MethodHierarchical Scene Coordinate Classification and Regression for Visual Localization
Li, X., Wang, S., Zhao, Y., Verbeek, J. J., & Kannala, J. (2020) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr42600.2020.01200 · Cited by 209
“Namely, either a random forest [4, 13, 14, 21, 30, 32, 33, 50, 57] or a neural network [3, 5, 6, 7, 8, 10, 11, 12, 27, 28, 30] is trained to directly predict 3D scene coordinates for the pixels and thus the 2D-3D correspondences are established.”
BackgroundAccelerated Coordinate Encoding: Learning to Relocalize in Minutes Using RGB and Poses
Brachmann, E., Cavallari, T., & Prisacariu, V. A. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.00488 · Cited by 202
“To regress correspondences, most works rely on random forests [6, 14, 15, 53, 61] or, more recently, convolutional neural networks [5,7,8,10,13,20,31].”
MethodReference Pose Generation for Long-term Visual Localization via Learned Features and View Synthesis
Zhang, Z., Sattler, T., & Scaramuzza, D. (2020) · International Journal of Computer Vision · doi:10.1007/s11263-020-01399-8 · Cited by 201
“2D-3D matching stage through a machine learning algorithm, typically either a random forest [5, 6, 85–91] or a CNN [90, 92–97].”
MethodCamera Relocalization by Computing Pairwise Relative Poses Using Convolutional Neural Network
Laskar, Z., Melekhov, I., Kalia, S., & Kannala, J. (2017) · 2017 IEEE International Conference on Computer Vision Workshops (ICCVW) · doi:10.1109/iccvw.2017.113 · Cited by 199
“[36] propose to exploit the uncertainty of the predicted 3D point locations during pose estimation.”
BackgroundMethodLearning to Navigate the Energy Landscape
Valentin, J., Dai, A., Niessner, M., Kohli, P., Torr, P., Izadi, S., & Keskin, C. (2016) · 2016 Fourth International Conference on 3D Vision (3DV) · doi:10.1109/3dv.2016.41 · Cited by 148
“Continuous Pose Refinement Similarly to [46, 40] which perform RGB-D relocalization by first generating hypothesis and then optimizing them in a continuous fashion for high precision, we run a continuous pose optimizer on top of the top prediction from our method.”
BackgroundMethodSelf-citationImage-Based Localization Using Hourglass Networks
Melekhov, I., Ylioinas, J., Kannala, J., & Rahtu, E. (2017) · 2017 IEEE International Conference on Computer Vision Workshops (ICCVW) · doi:10.1109/iccvw.2017.107 · Cited by 197
“As noted in [29], these methods suffer in situations where there are no strong constraints on the camera motion.”
BackgroundMethodInfluentialSelf-Supervised Visual Descriptor Learning for Dense Correspondence
Schmidt, T., Newcombe, R., & Fox, D. (2017) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2016.2634089 · Cited by 148
“[26] did, we compute for each pixel sample the set of all model”
BackgroundMethodInfluentialBehl, A., Jafari, O. H., Mustikovela, S. K., Alhaija, H. A., Rother, C., & Geiger, A. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.281 · Cited by 145
“For fine-grained geometric recognition we train a new convolutional neural network (CNN) on 2D instance segmentations of MNC which predicts the continuous 3D object parts, also known as 3D object coordinates [3,4,25,33, 34], for each pixel inside the instance mask.”
BackgroundMethodDelving deeper into convolutional neural networks for camera relocalization
Wu, J., Ma, L., & Hu, X. (2017) · 2017 IEEE International Conference on Robotics and Automation (ICRA) · doi:10.1109/icra.2017.7989663 · Cited by 133
“SCoRe Forests approach is extended with a probabilistic approach which exploits uncertainty from regression forests for pose estimation [4].”
BackgroundLaMAR: Benchmarking Localization and Mapping for Augmented Reality
Sarlin, P.-E., Dusmanu, M., Schönberger, J. L., Speciale, P., Gruber, L., Larsson, V., Mikšík, O., & Pollefeys, M. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-20071-7_40 · Cited by 166
Semantic Match Consistency for Long-Term Visual Localization
Toft, C., Stenborg, E., Hammarstrand, L., Brynte, L., Pollefeys, M., Sattler, T., & Kahl, F. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01216-8_24 · Cited by 165
“An alternative to obtaining 2D-3D correspondences via explicit feature matching is to directly learn the matching function [6, 7, 10, 33, 45, 50].”
BackgroundAre Large-Scale 3D Models Really Necessary for Accurate Visual Localization?
Torii, A., Taira, H., Sivic, J., Pollefeys, M., Okutomi, M., Pajdla, T., & Sattler, T. (2021) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2019.2941876 · Cited by 126
“Rather than replacing the complete localization pipeline, scene coordinate regression techniques [72], [73], [74], [75], [76], [77] only replace the 2D-3D matching stage while keeping the RANSAC-based pose estimation stage.”
MethodExpert Sample Consensus Applied to Camera Re-Localization
Brachmann, E., & Rother, C. (2019) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00762 · Cited by 155
“Some scene coordinate regression methods use an ensemble of base learners, namely random forests [44, 49, 7, 31, 32, 12, 33].”
BackgroundMethodOn-the-Fly Adaptation of Regression Forests for Online Camera Relocalisation
Cavallari, T., Golodetz, S., Lord, N. A., Valentin, J., Stefano, L. D., & Torr, P. H. S. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.31 · Cited by 121
“Spatial variance is defined in terms of the log of the determinant of the covariance of a fitted 3D Gaussian [37].”
BackgroundMethodInfluentialSelf-citationTo Learn or Not to Learn: Visual Localization from Essential Matrices
Zhou, Q., Sattler, T., Pollefeys, M., & Leal-Taixé, L. (2020) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra40945.2020.9196607 · Cited by 143
“Several works have been proposed to directly learn the 2D-3D matching function [5, 7, 14, 20, 24, 67, 76] While traditionally a dense 3D scene model is required for training, a recent work has shown that similar accuracy can be achieved without the need of a 3D model during training [7].”
BackgroundImage-based camera localization: an overview
Wu, Y., Tang, F., & Li, H. (2018) · Visual Computing for Industry Biomedicine and Art · doi:10.1186/s42492-018-0008-z · Cited by 132
SANet: Scene Agnostic Network for Camera Localization
Yang, L., Bai, Z., Tang, C., Li, H., Furukawa, Y., & Tan, P. (2019) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00013 · Cited by 127
“[31] trained a random forest to predict multi-model distributions of scene coordinates for increased pose accuracy.”
BackgroundAdvances in intelligent and autonomous navigation systems for small UAS
Bijjahalli, S., Sabatini, R., & Gardi, A. (2020) · Progress in Aerospace Sciences · doi:10.1016/j.paerosci.2020.100617 · Cited by 123
KFNet: Learning Temporal Camera Relocalization Using Kalman Filtering
Zhou, L., Luo, Z., Shen, T., Zhang, J., Zhen, M., Yao, Y., Fang, T., & Quan, L. (2020) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr42600.2020.00497 · Cited by 119
Large-scale, real-time visual–inertial localization revisited
Lynen, S., Zeisl, B., Aiger, D., Bosse, M. C., Hesch, J. A., Pollefeys, M., Siegwart, R., & Sattler, T. (2020) · The International Journal of Robotics Research · doi:10.1177/0278364920931151 · Cited by 114
“There are two dominant approaches in the literature: Learning to predict scene coordinates (Shotton et al. 2013; Valentin et al. 2015; Brachmann et al. 2017; Brachmann and Rother 2018; Cavallari et al. 2017) and learning to directly regress camera poses (Kendall et al.”
BackgroundReal-Time RGB-D Camera Pose Estimation in Novel Scenes Using a Relocalisation Cascade
Cavallari, T., Golodetz, S., Lord, N. A., Valentin, J., Prisacariu, V. A., Stefano, L. D., & Torr, P. H. S. (2019) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2019.2915068 · Cited by 112
“Initially, we train a regression forest offline to predict 2D-to-3D correspondences for a generic scene using the approach described in [27].”
BackgroundMethodInfluentialSelf-citationOn the Limits of Pseudo Ground Truth in Visual Camera Re-localisation
Brachmann, E., Humenberger, M., Rother, C., & Sattler, T. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00616 · Cited by 99
“Scene coordinate regression methods use a learned model, a neural network or a random forest, to predict dense image-to-scene correspondences [5–10, 15, 16, 42, 71, 82, 88].”
MethodRandom forests versus Neural Networks — What's best for camera localization?
Massiceti, D., Krull, A., Brachmann, E., Rother, C., & Torr, P. H. S. (2017) · 2017 IEEE International Conference on Robotics and Automation (ICRA) · doi:10.1109/icra.2017.7989598 · Cited by 73
“While they achieve moderate results in large-scale outdoor scenes, their performance in small, indoor scenes is one order of magnitude worse in pose translation error than [2] and [3].”
BackgroundMethodResultInfluentialSelf-citationGLACE: Global Local Accelerated Coordinate Encoding
Wang, F., Jiang, X., Galliani, S., Vogel, C., & Pollefeys, M. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.02037 · Cited by 82
“To regress 2D-3D matches, SCR methods are mainly based on random forests [9, 16, 18, 42, 51] or convolutional neural networks [6–8, 10, 12, 21, 26].”
Collaborative Large-Scale Dense 3D Reconstruction with Online Inter-Agent Pose Optimisation
Golodetz, S., Cavallari, T., Lord, N. A., Prisacariu, V. A., Murray, D. W., & Torr, P. H. S. (2018) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2018.2868533 · Cited by 78
“Moreover, in RGB-D relocalisation, keyframe-based relocalisers such as random ferns [23], which were previously widely used for relocalisation in a single-user context but were unable to relocalise from novel poses, have recently been giving way to high-quality regression-based methods such as SCoRe forests [49, 52],…”
MethodSelf-citationDeep Bingham Networks: Dealing with Uncertainty and Ambiguity in Pose Estimation
Deng, H., Bui, M., Navab, N., Guibas, L. J., Ilić, S., & Birdal, T. (2022) · International Journal of Computer Vision · doi:10.1007/s11263-022-01612-w · Cited by 76
“Valentin et al. (Valentin et al., 2015) stored components of a GaussianMixtureModel at the leaves of a scene coordinate regression forest (Shotton et al., 2013).”
BackgroundLi, M., Qin, Z., Gao, Z., Yi, R., Zhu, C., Guo, Y., & Xu, K. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.01299 · Cited by 74
“The main idea of them is to predict the 3D coordinates of each image pixel with decision trees [45, 34, 35, 4, 49] or neural networks [3, 5, 6, 28, 29, 33, 52].”
BackgroundText2Pos: Text-to-Point-Cloud Cross-Modal Localization
Kolmet, M., Zhou, Q., Ošep, A., & Leal-Taixé, L. (2022) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52688.2022.00657 · Cited by 71
“Related to our problem is the task of visual localization [4, 6, 16, 22, 32, 33, 35, 41, 45, 46, 56], which means estimating a precise pose based on an observed image or image sequence.”
BackgroundLearning to Detect Scene Landmarks for Camera Localization
Do, T., Mikšík, O., DeGol, J., Park, H. S., & Sinha, S. N. (2022) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52688.2022.01085 · Cited by 71
“SCR methods regress dense 3D scene coordinates to establish 2D-3D correspondences that are fed into pose solvers [35, 36, 41, 42, 69,82], allowing them to exploit geometric reasoning.”
BackgroundMethodVisual Camera Re-Localization Using Graph Neural Networks and Relative Pose Supervision
Türkoglu, M. Ö., Brachmann, E., Schindler, K., Brostow, G., & Monszpart, Á. (2021) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv53792.2021.00025 · Cited by 68
“Structure-based methods are thus far unchallenged in pose accuracy but, besides the computational demands of large-scale correspondence search, they require known camera intrinsics and a structured 3D representation of the scene, either from SfM [16, 52, 71, 40, 56, 51, 53], via implicit triangulation [8, 11] or by sc…”
BackgroundMethodVS-Net: Voting with Segmentation for Visual Localization
Huang, Z., Zhou, H., Li, Y., Yang, B., Xu, Y., Zhou, X., Bao, H., Zhang, G., & Li, H. (2021) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr46437.2021.00604 · Cited by 68
“Another method is to predict scene coordinates [7, 8, 49, 51, 11, 9, 25, 66] as an intermediate representation and estimate the camera pose through a RANSAC-PnP [18, 49] algorithm, which achieves state-of-the-art localization performance in small and medium scenes.”
MethodIndoor Visual Positioning Aided by CNN-Based Image Retrieval: Training-Free, 3D Modeling-Free
Chen, Y., Chen, R., Liu, M., Xiao, A., Wu, D., & Zhao, S. (2018) · Sensors · doi:10.3390/s18082692 · Cited by 67
Backtracking regression forests for accurate camera relocalization
Meng, L., Chen, J., Tung, F., Little, J. J., Valentin, J., & de Silva, C. W. (2017) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2017.8206611 · Cited by 64
“The random features are based on pairwise pixel comparison as in [1], [3]:”
MethodSelf-citationWang, K., Wang, Y., Liu, B., & Chen, J. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3256459 · Cited by 63
“Perceptual systems encounter significant obstacles in the actual world [38], [39], [40].”
BackgroundWang, S., Laskar, Z., Melekhov, I., Li, X., Zhao, Y., Tolias, G., & Kannala, J. (2024) · International Journal of Computer Vision · doi:10.1007/s11263-023-01982-9 · Cited by 61
“…et al., 2016; Cavallari et al., 2020, 2017; Guzmán-Rivera et al., 2014; Massiceti et al., 2017; Meng et al., 2017, 2018; Shotton et al., 2013; Valentin et al., 2015) or a neural network (Brachmann et al., 2017; Brachmann & Rother, 2018, 2019a,c, 2021; Budvytis et al., 2019; Bui et al., 2018;…”
Beyond Controlled Environments: 3D Camera Re-localization in Changing Indoor Scenes
Wald, J., Sattler, T., Golodetz, S., Cavallari, T., & Tombari, F. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58571-6_28 · Cited by 61
Cavallari, T., Bertinetto, L., Mukhoti, J., Torr, P. H. S., & Golodetz, S. (2019) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv.2019.00068 · Cited by 59
“(iv) Local regression methods generally use regression forests [63, 27, 69, 6, 48, 13, 49, 50, 12], neural networks [5, 7, 18, 42, 43, 8], or a mix of the two [46] to predict the scene coordinates of pixels in the input image.”
MethodSelf-citationVisual Localization via Few-Shot Scene Region Classification
Dong, S., Wang, S., Zhuang, Y., Kannala, J., Pollefeys, M., & Chen, B. (2022) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv57658.2022.00051 · Cited by 58
“There are also RGB-D based visual localization approaches [10, 11, 14, 39, 47] leveraging random forests or point cloud based backbones, with 3D-3D correspondence optimization algorithms, which is beyond our scope.”
BackgroundWang, Y., Zhang, S., Wan, B., He, W., & Bai, X. (2018) · The International Journal of Advanced Manufacturing Technology · doi:10.1007/s00170-018-2575-8 · Cited by 56
“…cloud-based tracking, researchers have also proposed the use of global pose estimation correction, including pose graph optimization [30], loop closure detection [31], incremental bundle adjustment [32, 33], or recovery from tracking failures by image-or keypoint-based relocalization [34, 35].”
BackgroundA Survey on Monocular Re-Localization: From the Perspective of Scene Map Representation
Miao, J., Jiang, K., Wen, T., Wang, Y., Jia, P., Wijaya, B., Zhao, X., Qian, C., Xiao, Z., Huang, J. H., et al. (2024) · IEEE Transactions on Intelligent Vehicles · doi:10.1109/tiv.2024.3378716 · Cited by 54
“Inthisfield, early approaches [222], [223] used regression forest models and they required RGB-D images as input.”
Visual Localization by Learning Objects-Of-Interest Dense Match Regression
Weinzaepfel, P., Csurka, G., Cabon, Y., & Humenberger, M. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.00578 · Cited by 51
“While random forests were used in the past [32, 38], Brachmann et al .”
MethodFull-Frame Scene Coordinate Regression for Image-Based Localization
Li, X., Ylioinas, J., & Kannala, J. (2018) · Robotics: Science and Systems · doi:10.15607/rss.2018.xiv.015 · Cited by 49
Image-based Localization with Spatial LSTMs.Preprint
Walch, F., Hazırbaş, C., Leal-Taixé, L., Sattler, T., Hilsenbeck, S., & Cremers, D. (2016) · arXiv (Cornell University) · Cited by 47
Safe and Robust Mobile Robot Navigation in Uneven Indoor Environments
Wang, C., Wang, J., Li, C., Ho, D., Cheng, J., Yan, T., Meng, L., & Meng, M. Q. ‐. (2019) · Sensors · doi:10.3390/s19132993 · Cited by 41
“The reader could refer to [24] to obtain more details about the random forests.”
Image-based localization using LSTMs for structured feature correlationPreprint
Walch, F., Hazırbaş, C., Leal-Taixé, L., Sattler, T., Hilsenbeck, S., & Cremers, D. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.07890 · Cited by 41
“explicitly model the uncertainty of the predicted 3D point positions and use this uncertainty during pose estimation [52], allowing them to localize more images .”
BackgroundMultiview Aggregation for Learning Category-Specific Shape ReconstructionPreprint
Sridhar, S., Rempe, D., Valentin, J., Bouaziz, S., & Guibas, L. J. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.01085 · Cited by 40
“This representation has been successfully used for several problems including camera pose estimation [37] and face reconstruction [10].”
MethodSelf-citationExploiting Points and Lines in Regression Forests for RGB-D Camera Relocalization
Meng, L., Tung, F., Little, J. J., Valentin, J., & de Silva, C. W. (2018) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2018.8593505 · Cited by 38
Self-citationVisual Localization using Imperfect 3D Models from the Internet
Panek, V., Kúkelová, Z., & Sattler, T. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.01266 · Cited by 39
R-SCoRe: Revisiting Scene Coordinate Regression for Robust Large-Scale Visual Localization
Jiang, X., Wang, F., Galliani, S., Vogel, C., & Pollefeys, M. (2025) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52734.2025.01077 · Cited by 37
“Following a similar structure-based localization paradigm as feature matching methods, SCR methods [4–9, 14, 15, 21, 36, 45, 60, 66, 68, 70] regress 2D-3D correspondences between the query image and the scene, and estimate the camera pose using geometric constraints.”
LaLaLoc++: Global Floor Plan Comprehension for Layout Localisation in Unvisited Environments
Howard-Jenkins, H., & Prisacariu, V. A. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19812-0_40 · Cited by 36
“Scene-coordinate regression methods [27,30,4,5,6] densely regress global 3D coordinates across image locations.”
SACReg: Scene-Agnostic Coordinate Regression for Visual Localization
Revaud, J., Cabon, Y., Brégier, R., Lee, J., & Weinzaepfel, P. (2024) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw63382.2024.00073 · Cited by 36
“Early approaches [25, 60, 72] used regression forest models to predict the correspondence of a pixel in a RGB-D frame to its 3D world coordinate.”
MethodDong, S., Fan, Q., Wang, H., Shi, J. F., Yi, L., Funkhouser, T. A., Chen, B., & Guibas, L. J. (2021) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr46437.2021.00844 · Cited by 36
“Following [54], we detect the modes of the empirical distribution in each leaf node via mean shift [14], and then construct a Gaussian Mixture Model (GMM) via iteratively estimating a 3D Gaussian distribution for each mode.”
BackgroundMethodContinual Learning for Image-Based Camera Localization
Wang, S., Laskar, Z., Melekhov, I., Li, X., & Kannala, J. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00324 · Cited by 36
“Scene coordinate regression methods [6–11,13 , 14 ,16 –18 ,26 ,37 –39 Reservoir sampling [66] samples a subset of data.”
F3Loc: Fusion and Filtering for Floorplan Localization
Chen, C., Wang, R., Vogel, C., & Pollefeys, M. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.01707 · Cited by 35
“Scene coordinate regression [6][42][44] learns to regress the 3D coordinates of the pixels in the query image.”
BackgroundSceneGraphLoc: Cross-Modal Coarse Visual Localization on 3D Scene Graphs
Miao, Y., Engelmann, F., Vysotska, O., Tombari, F., Pollefeys, M., & Baráth, D. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-73242-3_8 · Cited by 35
LaLaLoc: Latent Layout Localisation in Dynamic, Unvisited Environments
Howard-Jenkins, H., Ruiz-Sarmiento, J.-R., & Prisacariu, V. A. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00995 · Cited by 35
“Instead of regressing camera pose, scenecoordinate regression methods [29, 5, 6, 7, 34] densely predict 2D-3D correspondences between the query and the environment, which then allows for solving for pose via PnP.”
MethodScene Coordinate Regression with Angle-Based Reprojection Loss for Camera Relocalization
Li, X., Ylioinas, J., Verbeek, J. J., & Kannala, J. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-11015-4_19 · Cited by 35
A review of Visual-Based Localization
Xin, X., Jiang, J., & Zou, Y. (2019) · International Conference on Robotics, Intelligent Control and Artificial Intelligence (RICAI) · doi:10.1145/3366194.3366211 · Cited by 33
“[31] further introduced the multiscale navigation map into the original localization method to obtain the initial candidate pose.”
MethodWhen Regression Meets Manifold Learning for Object Recognition and Pose Estimation
Bui, M., Zakharov, S., Albarqouni, S., Ilić, S., & Navab, N. (2018) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2018.8460654 · Cited by 33
“Due to the recent successes of this work, it has been extended and optimized in several related methods [15], [16].”
MethodImage-based Localization using Hourglass NetworksPreprint
Melekhov, I., Ylioinas, J., Kannala, J., & Rahtu, E. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1703.07971 · Cited by 33
“As noted in [29], these methods suffer in situations where there are no strong constraints on the camera motion.”
BackgroundMethodInfluentialLearning to Localize in New Environments from Synthetic Training Data
Winkelbauer, D., Denninger, M., & Triebel, R. (2021) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra48506.2021.9560872 · Cited by 32
“There also exist learning-based approaches using CNNs [9]–[12] or random forests [9], [10], [13]–[15] which predict for every pixel of a given image the corresponding matching 3D coordinate.”
Background6D Camera Relocalization in Ambiguous Scenes via Continuous Multimodal Inference
Bui, M., Birdal, T., Deng, H., Albarqouni, S., Guibas, L. J., Ilić, S., & Navab, N. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58523-5_9 · Cited by 32
Learning Semantic-Aware Local Features for Long Term Visual Localization
Fan, B., Zhou, J., Feng, W., Pu, H., Yang, Y., Kong, Q., Wu, F., & Liu, H. (2022) · IEEE Transactions on Image Processing · doi:10.1109/tip.2022.3187565 · Cited by 32
Adversarial Networks for Camera Pose Regression and Refinement
Bui, M., Baur, C., Navab, N., Ilić, S., & Albarqouni, S. (2019) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2019.00470 · Cited by 31
“Most often correspondences are computed using for example SIFT [19] features or implicitly learned using regression forests [36, 41] as well as deep learning methods [5, 6, 8].”
MethodLiu, Y., Liu, L., Zheng, Y., Liu, Y., Dang, F., Li, N., & Ma, K. (2025) · Science China Information Sciences · doi:10.1007/s11432-024-4303-8 · Cited by 30
“Instead of using feature descriptors for matching, SCR directly feeds the query image to a random forest [142–144] or a neural network [145] to obtain the per-pixel scene coordinates of the query image.”
Huang, Z., Yu, H., Shentu, Y., Yuan, J., & Zhang, G. (2025) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52734.2025.02520 · Cited by 30
“SCR meth-ods do not directly predict camera pose but instead infer the dense scene coordinates [4, 15, 42, 47, 56, 58, 61].”
Leveraging Neural Radiance Fields for Uncertainty-Aware Visual Localization
Chen, L., Chen, W., Wang, R., & Pollefeys, M. (2024) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra57147.2024.10610126 · Cited by 30
Relative geometry-aware siamese neural network for 6DOF camera relocalization
Li, Q., Zhu, J., Cao, R., Sun, K., Garibaldi, J. M., Li, Q., Liu, B., & Qiu, G. (2020) · Neurocomputing · doi:10.1016/j.neucom.2020.09.071 · Cited by 30
FocusTune: Tuning Visual Localization through Focus-Guided Sampling
Nguyen, S. T., Fontan, A., Milford, M. J., & Fischer, T. (2024) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv57701.2024.00357 · Cited by 28
“APR models output the rotation (e.g. represented by a quaternion) and the translation vectors of the corresponding camera pose to the input image.”
BackgroundVNLSTM-PoseNet: A novel deep ConvNet for real-time 6-DOF camera relocalization in urban streets
Li, M., Qin, J., Li, D., Chen, R., Liao, X., & Guo, B. (2021) · Geo-spatial Information Science · doi:10.1080/10095020.2021.1960779 · Cited by 28
“Valentin et al. (2015) used the uncertainty in the model to further improve this method.”
MethodViPR: Visual-Odometry-aided Pose Regression for 6DoF Camera Localization
Ott, F., Feigl, T., Löffler, C., & Mutschler, C. (2020) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00029 · Cited by 27
“Hybrid methods [66, 7, 6, 23, 74] combine geometric and machine learning (ML) approaches.”
BackgroundMethodScene Coordinate and Correspondence Learning for Image-Based LocalizationPreprint
Bui, M., Albarqouni, S., Ilić, S., & Navab, N. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1805.08443 · Cited by 25
“Several extensions and improvements of the method have been proposed, increasing its performance and robustness [8], [6], [7].”
BackgroundMethodBuilding Fusion: Semantic-aware Structural Building-scale 3D Reconstruction
Zheng, T., Zhang, G., Han, L., Xu, L., & Fang, L. (2020) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2020.3042881 · Cited by 22
“Where am I?” Scene Retrieval with Language
Chen, J., Baráth, D., Armeni, I., Pollefeys, M., & Blum, H. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72913-3_12 · Cited by 21
ContextualNet: Exploiting Contextual Information Using LSTMs to Improve Image-Based Localization
Patel, M., Emery, B., & Chen, Y. (2018) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2018.8461124 · Cited by 21
CT-Loc: Cross-domain visual localization with a channel-wise transformer
Kim, D., & Kim, J. (2022) · Neural Networks · doi:10.1016/j.neunet.2022.11.014 · Cited by 19
SC-wLS: Towards Interpretable Feed-forward Camera Re-localization
Wu, X., Zhao, H., Li, S., Cao, Y., & Zha, H. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19769-7_34 · Cited by 19
“Apart from random forest based methods using RGB-D inputs [38,40,28], scene coordinate regression on RGB images is seeing steady progress [4,5,6,7,56].”
MethodDSAC - Differentiable RANSAC for Camera LocalizationPreprint
Brachmann, E., Krull, A., Nowozin, S., Shotton, J., Michel, F., Gumhold, S., & Rother, C. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.05705 · Cited by 19
“We demonstrate an end-to-end trainable solution for this problem, building on the scene coordinate regression forest (SCoRF) approach [36, 40, 5].”
MethodSelf-citationEvaluation Criteria for Inside-Out Indoor Positioning Systems Based on Machine Learning
Löffler, C., Riechel, S., Fischer, J., & Mutschler, C. (2018) · International Conference on Indoor Positioning and Indoor Navigation · doi:10.1109/ipin.2018.8533862 · Cited by 18
ACE-G: Improving Generalization of Scene Coordinate Regression Through Query Pre-Training
Bruns, L., Barroso-Laguna, A., Cavallari, T., Monszpart, Á., Munukutla, S., Prisacariu, V. A., & Brachmann, E. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.02483 · Cited by 17
“Initially proposed for RGB-D inputs (using random forests [18, 53, 57]), SCR methods have later been adopted for RGB-only inputs, using neural networks [9, 11, 12, 36, 37, 40, 62].”
Metadata
- Authors
- Valentin, J., Niebner, M., Shotton, J., Fitzgibbon, A., Izadi, S., & Torr, P.
- Affiliations
- University of Oxford (GB)
- Stanford University (US)
- Microsoft (United States) (US)
- Venue
- 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
- Publisher
- IEEE
- Published
- 2015-06
- Pages
- 4400-4408
- ISSN
- 1063-6919 (ISSN-L)
- OpenAlex
- W1893935112
- Cited by
- 156
- References
- 23
- Citations by year
- 2026: 4 · 2025: 9 · 2024: 13 · 2023: 7 · 2022: 13 · 2021: 19 · 2020: 19 · 2019: 25 · 2018: 20 · 2017: 15 · 2016: 12
- FWCI
- 27.16
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings (publishedVersion)
- Funders
- Leverhulme Trust
- Engineering and Physical Sciences Research Council (EP/I001107/2)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Robotics and Sensor-Based Localization
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Vision and Imaging
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image and Video Retrieval Techniques
- Keywords
- camera relocalization, regression forests, uncertainty estimation, camera pose optimization, visual SLAM
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.