Pith. sign in

REVIEW 4 major objections 4 minor 58 references

RA-NeRF: Robust Neural Radiance Field Reconstruction with Accurate Camera Pose Estimation under Complex Trajectories

T0 review · 4 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read RA-NeRF recovers camera poses from raw image sequences with about one-degree mean rotation error, even under complex trajectories.

desk verdict A plausible incremental pose-free NeRF pipeline with strong ablations, but a suspect essential-matrix formula and missing closest baselines keep the SOTA claim from being fully established. read the letter →

arxiv 2506.15242 v2 pith:HWAL2X6P submitted 2025-06-18 cs.CV

Qingsong Yan , Qiang Wang , Kaiyong Zhao , Jie Chen , Bo Li , Xiaowen Chu , Fei Deng This is my paper · ORCID
classification cs.CV
keywords neuralradiancefieldscameraposeestimationopticalflowregularizationincrementallearningnovelviewsynthesisSE(3)updatesimplicitfilter
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

RA-NeRF claims that camera poses accurate enough for high-quality 3D reconstruction can be recovered directly from raw image sequences, even when the camera follows large rotational trajectories rather than translating forward. The key claim is that a learned optical-flow network can supply reliable relative-pose supervision during incremental NeRF optimization, while an implicit pose filter denoises the pose updates. On a benchmark of rotational camera trajectories the method reports a mean rotation error of about 1.14 degrees and translation error of about 0.54, down from roughly 7 to 11 degrees for earlier pose-free pipelines, with state-of-the-art rendering quality. If the claims hold, NeRF and 3D Gaussian Splatting pipelines would no longer need a separate structure-from-motion stage to supply camera poses before reconstruction.

What carries the argument

Flow-driven pose regulation is the central object: for a pair of images, sampled correspondences from a learned optical-flow network go through the eight-point algorithm to give a fundamental matrix, which is converted to an essential matrix and decomposed by SVD into four candidate relative poses; the one with positive triangulated depth is chosen. The loss term then measures the distance between that flow-derived relative pose and the relative pose formed from the current camera estimates, with translation compared only in direction. The second mechanism is an implicit pose filter, an 8-layer MLP with a residual connection that takes a shared global motion embedding and a per-image local embedding and outputs a six-dimensional update per image in the Lie algebra se(3), converted to a rigid transformation by the exponential map. Camera poses are updated on the Lie group SE(3), the group of rigid camera motions, via the exponential map, which the ablations identify as more stable than adding a raw se(3) vector.

What would settle it

Run RA-NeRF on a trajectory with ground-truth poses and with a scene containing large textureless or reflective regions where learned optical flow is known to fail; then disable the flow-driven regulation and compare. If pose accuracy does not degrade, or if the regularization visibly increases error, the flow supervision is not the load-bearing component the paper claims it is. A more direct test: corrupt the optical flow during the initialization stage and check whether the final pose error mirrors the corruption.

Watch

Extended reading notes

Core claim

RA-NeRF establishes that the relative pose between two images, computed directly from optical-flow correspondences by the eight-point algorithm and essential-matrix decomposition, can act as an external supervisory signal that keeps NeRF pose optimization from sinking into photometric local minima. The method compares these flow-derived relative rotations and translation directions with the relative poses implied by the current camera parameters and penalizes the difference, supervising only translation direction because the flow-derived translation has no absolute scale. It further replaces raw per-pose learnable updates with an implicit pose filter, a residual network that reads a global motion embedding and per-image local embeddings and outputs an se(3) update, then applies the update through the exponential map. The reported results on two real-world benchmarks support the claim that this combination yields state-of-the-art camera-pose accuracy and novel-view quality from unposed images.

Load-bearing premise

The load-bearing premise is that optical-flow correspondences, after positive-depth disambiguation, give a trustworthy relative-pose signal even when the current camera estimates are far off; if the flow is wrong or the disambiguation picks the wrong solution, the regularizer actively drags poses away from the truth.

Editorial extensions

If this is right

  • Pose-free NeRF reconstruction becomes practical for large-rotation trajectories, not only forward-looking scenes.
  • Incremental pipelines can use flow-derived relative poses instead of depth priors or known pose distributions to avoid local minima.
  • Supervising only the direction of translation sidesteps the scale ambiguity of monocular optical flow, keeping the regularization usable from scratch.
  • The implicit pose filter shows that learned denoising of pose gradients can be shared across frames through a global motion embedding, improving pose accuracy in complex motions.
  • The full pipeline offers an end-to-end route from unposed images to a reconstructed NeRF, removing a separate structure-from-motion preprocessing stage.

Reading between the lines

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

  • The regularization's dependence on optical flow suggests the method should degrade on scenes with reflections, repetitive texture, or near-degenerate camera motion; the paper does not test these failure modes.
  • Because only translation direction is supervised, the absolute scale of the trajectory is pinned down by the NeRF photometric optimization; a testable prediction is that low-parallax videos will show larger translation drift than rotation error.
  • The same flow-regulation and implicit-filter components could in principle be attached to 3D Gaussian Splatting or a hash-grid NeRF backbone; the paper only demonstrates them with a plain MLP NeRF.
  • The global motion embedding may encode scene-specific motion patterns, so a filter trained on one scene may need re-initialization when applied to a different trajectory; adapting it across scenes without re-training remains an open question.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper proposes RA-NeRF, an incremental NeRF-based pipeline that estimates camera poses and reconstructs a scene jointly from unposed image sequences. Two technical components are introduced: a flow-driven pose regulation term that derives relative pose targets from PDC-Net optical flow via the eight-point method and essential-matrix decomposition, and an implicit pose filter, an MLP that maps global and local motion embeddings to per-image SE(3) updates. The method is evaluated on the NeRFBuster and Tanks&Temple datasets, where it reports substantially lower rotation and translation errors than CF-NeRF, CF-3DGS, Nope-NeRF, BARF, and related baselines, along with better PSNR and LPIPS. Ablations in Tables III, IV, and V isolate the contributions of the flow regulation term and the implicit pose filter.

Significance. If the reported results are reproducible, the paper makes a strong practical contribution: it demonstrates that an incremental, photometric-plus-flow approach can estimate camera poses from scratch on rotation-heavy trajectories, a regime where global bundle-adjusting NeRF methods fail badly. The central mechanism is not circular, since the flow-derived relative poses come from a pretrained external network (PDC-Net) and not from the optimized pose variables. The ablations are informative and show large performance changes when either proposed component is removed, and the pose errors reported for NeRFBuster are an order of magnitude below the closest incremental baseline. The main risks are the reliability of the flow-derived supervision targets and the dataset-specific tuning of the flow loss weight, both of which are load-bearing for the method's central claim.

major comments (4)
  1. [Section III-B, Eq. (2)] The conversion from the fundamental matrix to the essential matrix is stated as E_ij = K^{-1}_j F_ij K^{-1}_i. The standard conversion is E = K^T F K (with the appropriate ordering for two cameras). As written, the computation does not yield an essential matrix, so the relative pose (R_hat, T_hat) obtained by SVD would not be geometrically valid. Because this target directly supervises the poses in Eq. (2), this is a load-bearing technical point that must be corrected or explained with a clear convention for the K matrices.
  2. [Section III-B, Eq. (2)] The paper never measures how often the flow-derived relative pose (R_hat_ij, T_hat_ij) is correct on the actual pairs used (adjacent frames plus N_loc=2 neighbors) in NeRFBuster and Tanks&Temple. The ablation in Table III shows that this term is indispensable, but if PDC-Net flow or the positive-depth disambiguation fails on small-baseline or near-pure-rotation pairs, the regulation in Eq. (2) actively pulls poses toward an incorrect target. The authors should report a success-rate or error histogram of the flow-derived poses against COLMAP ground truth on the evaluation pairs, and discuss how failures are handled.
  3. [Section IV-B, Tables III and IV] The flow regulation weight lambda_f is set to 1e-1 on NeRFBuster and 1e-3 on Tanks&Temple, a change of two orders of magnitude between datasets, yet no sensitivity analysis or multiple-seed evaluation is reported. This weakens the robustness claim in the conclusion: it remains possible that the method's success depends on carefully selecting this weight per dataset. A sensitivity table (e.g., lambda_f in {1e-4, 1e-3, 1e-2, 1e-1}) and repeated runs with different seeds would establish the stability of the reported mean errors.
  4. [Section IV-D, Tables I and II] The related work cites CT-NeRF and LocalRF as the closest flow-based and incremental baselines, but neither appears in the quantitative comparison. Since the paper claims state-of-the-art performance on both datasets, omitting these published baselines makes the comparison incomplete and overstates the SOTA claim. The authors should add these methods or justify their absence with a concrete explanation (e.g., code or setting incompatibility).
minor comments (4)
  1. [Section IV-E-b] The text says "IF-NeRF proposes the implicit pose filter," but IF-NeRF is not introduced anywhere; this appears to be a typo for RA-NeRF and should be corrected.
  2. [References] Reference [16] contains a typo in the title: "imporved-nerfmm" should be "improved-nerfmm."
  3. [Introduction and Section III-D] The phrase "high-accurate" should be "high-accuracy," and the definition of lambda_i in Eq. (4) is ambiguous: Section III-D states lambda_i=1 during initialization/localization and lambda_i=0 during optimization, but the role of lambda_i versus lambda_f should be stated explicitly when the loss is first introduced.
  4. [Table V] The table header uses "se(3)" and "SE(3)" with inconsistent capitalization; please unify the notation to SE(3) for the manifold and se(3) for the Lie algebra.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the flow-driven supervision is external to the optimized poses and the main results are evaluated against COLMAP ground truth on independent benchmarks.

full rationale

The central mechanism of RA-NeRF is the flow-driven pose regulation in Eq. (2), which supervises the estimated relative poses (R_ij, T_ij) with the relative poses (R_hat_ij, T_hat_ij) recovered from PDC-Net optical flow via the eight-point method and essential-matrix decomposition. These flow-derived targets are computed from the input images independently of the current pose estimates, so the loss is not defined in terms of the quantity being predicted. The implicit pose filter is a learned MLP that outputs pose updates, but it is trained jointly with the same photometric and flow losses; it does not encode the final poses as inputs. The evaluation uses external datasets (NeRFBuster and Tanks&Temple) with COLMAP ground truth, and the ablation studies demonstrate the contribution of each component without relying on self-citation. The paper does cite the authors' own CF-NeRF as the base incremental pipeline, but this is prior published work used as a starting point, not as a proof of the new claim. The claimed improvements are supported by quantitative comparisons against external baselines. There is no step where a prediction is equivalent to a fitted input, no uniqueness theorem imported from the authors' prior work, and no ansatz smuggled via citation. The potential sensitivity to the reliability of PDC-Net flow on small-baseline or rotation-heavy pairs is a correctness concern, not a circularity, and the E = K^{-1} F K^{-1} notation appears to be a typographical or formula issue rather than a circular construct. Overall, the derivation chain is self-contained and the central claims are not forced by definition or by self-citation.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

No new physical entities such as particles, forces, dimensions, or conserved quantities are introduced. The global and local motion embeddings and the implicit pose filter are learned parameters and network modules, not postulated entities with independent falsifiable handles outside the paper.

free parameters (7)
  • lambda_f = 1e-1 (NeRFBuster) and 1e-3 (Tanks and Temple)
    Weight of the flow-driven pose regulation loss in Eq. 4; set differently per dataset, making the claimed advantage potentially dependent on per-dataset tuning.
  • lambda_r = 3
    Weight on the rotation difference term in Eq. 2; chosen by hand.
  • lambda_i = 1 in initialization and localization, 0 in optimization
    Hard-coded gate that turns flow-driven pose regulation on only during initialization and localization stages.
  • iteration counts = xi_init = xi_part = xi_glob = 3100, xi_loca = 1100
    Number of training iterations per incremental stage; chosen by hand and reported in Section IV-B.
  • window sizes = N_init = N_loc = 2, N_part = N_glob = 5
    Number of images used in each incremental stage; chosen by hand.
  • initial learning rate = 5e-4
    Adam learning rate used for both datasets and all stages.
  • motion embedding dimension = 256
    Dimension of the global and local motion embeddings that feed the implicit pose filter.
assumptions (6)
  • domain assumption Known camera intrinsics K are available for the essential-matrix conversion E = K^{-T} F K^{-1}.
    The pose-from-flow constraint in Section III-B assumes calibrated cameras; unposed casual captures may have unknown or varying intrinsics.
  • domain assumption Optical flow from PDC-Net is reliable enough to yield correct relative poses via the eight-point algorithm.
    The flow-driven pose regulation depends on accurate correspondences; the paper provides no failure analysis for textureless regions or low-overlap pairs.
  • domain assumption The positive-depth triangulation rule resolves the four-fold ambiguity of essential-matrix decomposition correctly in all training pairs.
    Section III-B selects one of four possible relative poses; if the selection is wrong, Eq. 2 pushes the optimized poses toward an incorrect relative pose.
  • standard math Eight-point algorithm and essential-matrix SVD decomposition are valid under pinhole multi-view geometry.
    Standard geometric tools from Hartley and Zisserman, used in Section III-B without further proof.
  • domain assumption COLMAP poses are accurate enough to serve as ground truth on NeRFBuster complex rotational trajectories.
    Evaluation in Section IV-C and the visual comparison in Fig. 1 use COLMAP poses as reference, yet the paper targets trajectories where COLMAP-style SfM is known to be fragile.
  • domain assumption An 8-layer MLP with residual connections can learn a motion pattern that suppresses gradient noise.
    The denoising capability of the implicit pose filter is asserted rather than derived; only ablative empirical evidence is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of RA-NeRF: Robust Neural Radiance Field Reconstruction with Accurate Camera Pose Estimation under Complex Trajectories." pith.science (2026). https://pith.science/paper/HWAL2X6P

@misc{pith2026250615242,
  author       = {Pith},
  title        = {Pith review of: RA-NeRF: Robust Neural Radiance Field Reconstruction with Accurate Camera Pose Estimation under Complex Trajectories},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HWAL2X6P}},
  note         = {Machine review of arXiv:2506.15242}
}
read the original abstract

Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS) have emerged as powerful tools for 3D reconstruction and SLAM tasks. However, their performance depends heavily on accurate camera pose priors. Existing approaches attempt to address this issue by introducing external constraints but fall short of achieving satisfactory accuracy, particularly when camera trajectories are complex. In this paper, we propose a novel method, RA-NeRF, capable of predicting highly accurate camera poses even with complex camera trajectories. Following the incremental pipeline, RA-NeRF reconstructs the scene using NeRF with photometric consistency and incorporates flow-driven pose regulation to enhance robustness during initialization and localization. Additionally, RA-NeRF employs an implicit pose filter to capture the camera movement pattern and eliminate the noise for pose estimation. To validate our method, we conduct extensive experiments on the Tanks\&Temple dataset for standard evaluation, as well as the NeRFBuster dataset, which presents challenging camera pose trajectories. On both datasets, RA-NeRF achieves state-of-the-art results in both camera pose estimation and visual quality, demonstrating its effectiveness and robustness in scene reconstruction under complex pose trajectories.

Figures

Figures reproduced from arXiv: 2506.15242 by the authors.

Figure 1
Figure 1. Visualization of estimated camera poses for one scene [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The pipeline of the RA-NeRF, which uses photometric consistency to train the NeRF and camera poses, and improves [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Given the essential matrix Eij of matched pixels pi , pj , we can decompose Eij with four possible poses through multi-view geometry [54], which can mislead pro￾jected ray distance [23] and Sampson distance [53]. randomly selects a subset of the matching pixels in each iteration to compute the relative camera pose, using a strategy similar to RANSAC to improve efficiency and robustness. C. Implicit Pose Filter Based… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Visualizations of estimated camera poses and rendering results. (a)-(c) are results from NeRFBuster [4] and (f)-(j) [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

58 extracted references · 26 canonical work pages

  1. [1]

    Nope-nerf: Optimising neural radiance field with no pose prior,

    W. Bian, Z. Wang, K. Li, J.-W. Bian, and V . A. Prisacariu, “Nope-nerf: Optimising neural radiance field with no pose prior,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition, 2023, pp. 4160–4169

    work page 2023

  2. [2]

    Cf- nerf: Camera parameter free neural radiance fields with incremental learning,

    Q. Yan, Q. Wang, K. Zhao, J. Chen, B. Li, X. Chu, and F. Deng, “Cf- nerf: Camera parameter free neural radiance fields with incremental learning,” inProceedings of the AAAI Conference on Artificial Intel- ligence, vol. 38, no. 6, 2024, pp. 6440–6448

    work page 2024

  3. [3]

    Colmap-free 3d gaussian splatting,

    Y . Fu, S. Liu, A. Kulkarni, J. Kautz, A. A. Efros, and X. Wang, “Colmap-free 3d gaussian splatting,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2024, pp. 20 796–20 805

    2024

  4. [4]

    Nerfbusters: Removing ghostly artifacts from casually captured nerfs,

    F. Warburg, E. Weber, M. Tancik, A. Holynski, and A. Kanazawa, “Nerfbusters: Removing ghostly artifacts from casually captured nerfs,”arXiv preprint arXiv:2304.10532, 2023

    arXiv 2023

  5. [5]

    Nerf: Representing scenes as neural radiance fields for view synthesis,

    B. Mildenhall, P. P. Srinivasan, M. Tancik, J. T. Barron, R. Ramamoor- thi, and R. Ng, “Nerf: Representing scenes as neural radiance fields for view synthesis,” inComputer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part I, 2020, pp. 405–421

    work page 2020

  6. [6]

    3d gaussian splatting for real-time radiance field rendering,

    B. Kerbl, G. Kopanas, T. Leimk ¨uhler, and G. Drettakis, “3d gaussian splatting for real-time radiance field rendering,”ACM Transactions on Graphics, vol. 42, no. 4, July 2023

    2023

  7. [7]

    D- nerf: Neural radiance fields for dynamic scenes,

    A. Pumarola, E. Corona, G. Pons-Moll, and F. Moreno-Noguer, “D- nerf: Neural radiance fields for dynamic scenes,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 10 318–10 327

    2021

  8. [8]

    Dynibar: Neural dynamic image-based rendering,

    Z. Li, Q. Wang, F. Cole, R. Tucker, and N. Snavely, “Dynibar: Neural dynamic image-based rendering,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023

    work page 2023

Show all 58 references
  1. [9]

    Neuralangelo: High-fidelity neural surface reconstruction,

    Z. Li, T. M ¨uller, A. Evans, R. H. Taylor, M. Unberath, M.-Y . Liu, and C.-H. Lin, “Neuralangelo: High-fidelity neural surface reconstruction,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 8456–8465

    2023

  2. [10]

    Pgsr: Planar-based gaussian splatting for efficient and high-fidelity surface reconstruction,

    D. Chen, H. Li, W. Ye, Y . Wang, W. Xie, S. Zhai, N. Wang, H. Liu, H. Bao, and G. Zhang, “Pgsr: Planar-based gaussian splatting for efficient and high-fidelity surface reconstruction,”arXiv preprint arXiv:2406.06521, 2024

    2024 arXiv

  3. [11]

    Switch-nerf: Learning scene decomposition with mixture of experts for large-scale neural radiance fields,

    M. Zhenxing and D. Xu, “Switch-nerf: Learning scene decomposition with mixture of experts for large-scale neural radiance fields,” in International Conference on Learning Representations, 2023

    2023

  4. [12]

    Vastgaussian: Vast 3d gaussians for large scene reconstruction,

    J. Lin, Z. Li, X. Tang, J. Liu, S. Liu, J. Liu, Y . Lu, X. Wu, S. Xu, Y . Yan,et al., “Vastgaussian: Vast 3d gaussians for large scene reconstruction,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 5166–5175

    2024

  5. [13]

    Structure-from-motion revisited,

    J. L. Schonberger and J.-M. Frahm, “Structure-from-motion revisited,” inProceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 4104–4113

    2016

  6. [14]

    Nerf– : Neural radiance fields without known camera parameters,

    Z. Wang, S. Wu, W. Xie, M. Chen, and V . A. Prisacariu, “Nerf– : Neural radiance fields without known camera parameters,”arXiv preprint arXiv:2102.07064, 2021

    2021 arXiv

  7. [15]

    Barf: Bundle- adjusting neural radiance fields,

    C.-H. Lin, W.-C. Ma, A. Torralba, and S. Lucey, “Barf: Bundle- adjusting neural radiance fields,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 5741–5751

    2021

  8. [16]

    imporved-nerfmm,

    J. Guo and A. Sherwood, “imporved-nerfmm,” https://github.com/ventusff/improved-nerfmm, 2021

    2021

  9. [17]

    Gaussian acti- vated neural radiance fields for high fidelity reconstruction and pose estimation,

    S.-F. Chng, S. Ramasinghe, J. Sherrah, and S. Lucey, “Gaussian acti- vated neural radiance fields for high fidelity reconstruction and pose estimation,” inEuropean Conference on Computer Vision. Springer, 2022, pp. 264–280

    2022

  10. [18]

    Local-to-global registration for bundle-adjusting neural radiance fields,

    Y . Chen, X. Chen, X. Wang, Q. Zhang, Y . Guo, Y . Shan, and F. Wang, “Local-to-global registration for bundle-adjusting neural radiance fields,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 8264–8273

    2023

  11. [19]

    Lu-nerf: Scene and pose estimation by synchronizing local unposed nerfs,

    Z. Cheng, C. Esteves, V . Jampani, A. Kar, S. Maji, and A. Makadia, “Lu-nerf: Scene and pose estimation by synchronizing local unposed nerfs,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 18 312–18 321

    2023

  12. [20]

    Local light field fusion: Practical view synthesis with prescriptive sampling guidelines,

    B. Mildenhall, P. P. Srinivasan, R. Ortiz-Cayon, N. K. Kalantari, R. Ramamoorthi, R. Ng, and A. Kar, “Local light field fusion: Practical view synthesis with prescriptive sampling guidelines,”ACM Transactions on Graphics (TOG), 2019

    2019

  13. [21]

    Td-nerf: Novel truncated depth prior for joint camera pose and neural radiance field optimization,

    Z. Tan, Z. Zhou, Y . Ge, Z. Wang, X. Chen, and D. Hu, “Td-nerf: Novel truncated depth prior for joint camera pose and neural radiance field optimization,” in2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2024, pp. 372–379

    2024

  14. [22]

    Vision transformers for dense prediction,

    R. Ranftl, A. Bochkovskiy, and V . Koltun, “Vision transformers for dense prediction,” inProceedings of the IEEE/CVF international conference on computer vision, 2021, pp. 12 179–12 188

    2021

  15. [23]

    Self-calibrating neural radiance fields,

    Y . Jeong, S. Ahn, C. Choy, A. Anandkumar, M. Cho, and J. Park, “Self-calibrating neural radiance fields,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 5846–5854

    2021

  16. [24]

    Porf: Pose resid- ual field for accurate neural surface reconstruction,

    J.-W. Bian, W. Bian, V . A. Prisacariu, and P. Torr, “Porf: Pose resid- ual field for accurate neural surface reconstruction,”arXiv preprint arXiv:2310.07449, 2023

    2023 arXiv

  17. [25]

    Sparf: Neural radiance fields from sparse and noisy poses,

    P. Truong, M.-J. Rakotosaona, F. Manhardt, and F. Tombari, “Sparf: Neural radiance fields from sparse and noisy poses,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition, 2023, pp. 4190–4200

    2023

  18. [26]

    Learning accu- rate dense correspondences and when to trust them,

    P. Truong, M. Danelljan, L. Van Gool, and R. Timofte, “Learning accu- rate dense correspondences and when to trust them,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 5714–5724

    2021

  19. [27]

    Loftr: Detector- free local feature matching with transformers,

    J. Sun, Z. Shen, Y . Wang, H. Bao, and X. Zhou, “Loftr: Detector- free local feature matching with transformers,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 8922–8931

    2021

  20. [28]

    Gnerf: Gan-based neural radiance field without posed camera,

    Q. Meng, A. Chen, H. Luo, M. Wu, H. Su, L. Xu, X. He, and J. Yu, “Gnerf: Gan-based neural radiance field without posed camera,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 6351–6361

    2021

  21. [29]

    Pose-free neural radiance fields via implicit pose regularization,

    J. Zhang, F. Zhan, Y . Yu, K. Liu, R. Wu, X. Zhang, L. Shao, and S. Lu, “Pose-free neural radiance fields via implicit pose regularization,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 3534–3543

    2023

  22. [30]

    Dbarf: Deep bundle-adjusting generalizable neural radiance fields,

    Y . Chen and G. H. Lee, “Dbarf: Deep bundle-adjusting generalizable neural radiance fields,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 24–34

    2023

  23. [31]

    Droid-slam: Deep visual slam for monocular, stereo, and rgb-d cameras,

    Z. Teed and J. Deng, “Droid-slam: Deep visual slam for monocular, stereo, and rgb-d cameras,”Advances in neural information processing systems, vol. 34, pp. 16 558–16 569, 2021

    2021

  24. [32]

    Progressively optimized local radiance fields for robust view synthesis,

    A. Meuleman, Y .-L. Liu, C. Gao, J.-B. Huang, C. Kim, M. H. Kim, and J. Kopf, “Progressively optimized local radiance fields for robust view synthesis,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 16 539–16 548

    2023

  25. [33]

    Ct-nerf: Incremental optimizing neural radiance field and poses with complex trajectory,

    Y . Ran, Y . Li, Q. Ye, Y . Huo, Z. Bai, J. Sun, and J. Chen, “Ct-nerf: Incremental optimizing neural radiance field and poses with complex trajectory,”arXiv preprint arXiv:2404.13896, 2024

    2024 arXiv

  26. [34]

    Dkm: Dense kernelized feature matching for geometry estimation,

    J. Edstedt, I. Athanasiadis, M. Wadenb ¨ack, and M. Felsberg, “Dkm: Dense kernelized feature matching for geometry estimation,” inPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 17 765–17 775

    2023

  27. [35]

    Tanks and temples: Benchmarking large-scale scene reconstruction,

    A. Knapitsch, J. Park, Q.-Y . Zhou, and V . Koltun, “Tanks and temples: Benchmarking large-scale scene reconstruction,”ACM Transactions on Graphics (ToG), vol. 36, no. 4, pp. 1–13, 2017

    2017

  28. [36]

    Plenoctrees for real-time rendering of neural radiance fields,

    A. Yu, R. Li, M. Tancik, H. Li, R. Ng, and A. Kanazawa, “Plenoctrees for real-time rendering of neural radiance fields,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 5752–5761

    2021

  29. [37]

    Instant neural graphics primitives with a multiresolution hash encoding,

    T. M ¨uller, A. Evans, C. Schied, and A. Keller, “Instant neural graphics primitives with a multiresolution hash encoding,”ACM Transactions on Graphics (ToG), vol. 41, no. 4, pp. 1–15, 2022

    2022

  30. [38]

    Mip-nerf 360: Unbounded anti-aliased neural radiance fields,

    J. T. Barron, B. Mildenhall, D. Verbin, P. P. Srinivasan, and P. Hedman, “Mip-nerf 360: Unbounded anti-aliased neural radiance fields,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 5470–5479

    2022

  31. [39]

    Fregs: 3d gaussian splatting with progressive frequency regularization,

    J. Zhang, F. Zhan, M. Xu, S. Lu, and E. Xing, “Fregs: 3d gaussian splatting with progressive frequency regularization,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion, 2024, pp. 21 424–21 433

    2024

  32. [40]

    Lightgaus- sian: Unbounded 3d gaussian compression with 15x reduction and 200+ fps,

    Z. Fan, K. Wang, K. Wen, Z. Zhu, D. Xu, and Z. Wang, “Lightgaus- sian: Unbounded 3d gaussian compression with 15x reduction and 200+ fps,”arXiv preprint arXiv:2311.17245, 2023

    2023 arXiv

  33. [41]

    Mip-nerf: A multiscale representation for anti- aliasing neural radiance fields,

    J. T. Barron, B. Mildenhall, M. Tancik, P. Hedman, R. Martin-Brualla, and P. P. Srinivasan, “Mip-nerf: A multiscale representation for anti- aliasing neural radiance fields,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 5855–5864

    2021

  34. [42]

    Mip-splatting: Alias-free 3d gaussian splatting,

    Z. Yu, A. Chen, B. Huang, T. Sattler, and A. Geiger, “Mip-splatting: Alias-free 3d gaussian splatting,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 19 447–19 456

    2024

  35. [43]

    Putting nerf on a diet: Semantically consistent few-shot view synthesis,

    A. Jain, M. Tancik, and P. Abbeel, “Putting nerf on a diet: Semantically consistent few-shot view synthesis,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 5885–5894

    2021

  36. [44]

    Freenerf: Improving few-shot neural rendering with free frequency regularization,

    J. Yang, M. Pavone, and Y . Wang, “Freenerf: Improving few-shot neural rendering with free frequency regularization,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 8254–8263

    2023

  37. [45]

    V olume rendering of neural implicit surfaces,

    L. Yariv, J. Gu, Y . Kasten, and Y . Lipman, “V olume rendering of neural implicit surfaces,”Advances in Neural Information Processing Systems, vol. 34, pp. 4805–4815, 2021

    2021

  38. [46]

    Neus: Learning neural implicit surfaces by volume rendering for multi-view reconstruction,

    P. Wang, L. Liu, Y . Liu, C. Theobalt, T. Komura, and W. Wang, “Neus: Learning neural implicit surfaces by volume rendering for multi-view reconstruction,”arXiv preprint arXiv:2106.10689, 2021

    2021 arXiv

  39. [47]

    Sugar: Surface-aligned gaussian splatting for efficient 3d mesh reconstruction and high-quality mesh rendering,

    A. Gu ´edon and V . Lepetit, “Sugar: Surface-aligned gaussian splatting for efficient 3d mesh reconstruction and high-quality mesh rendering,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 5354–5363

    2024

  40. [48]

    Geo-neus: Geometry-consistent neural implicit surfaces learning for multi-view reconstruction,

    Q. Fu, Q. Xu, Y . S. Ong, and W. Tao, “Geo-neus: Geometry-consistent neural implicit surfaces learning for multi-view reconstruction,”Ad- vances in Neural Information Processing Systems, vol. 35, pp. 3403– 3416, 2022

    2022

  41. [49]

    inerf: Inverting neural radiance fields for pose estimation,

    L. Yen-Chen, P. Florence, J. T. Barron, A. Rodriguez, P. Isola, and T.-Y . Lin, “inerf: Inverting neural radiance fields for pose estimation,” in2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2021, pp. 1323–1330

    2021

  42. [50]

    Samurai: Shape and material from unconstrained real-world arbitrary image collections,

    M. Boss, A. Engelhardt, A. Kar, Y . Li, D. Sun, J. T. Barron, H. Lensch, and V . Jampani, “Samurai: Shape and material from unconstrained real-world arbitrary image collections,”arXiv preprint arXiv:2205.15768, 2022

    2022 arXiv

  43. [51]

    Mononerf: Learning a generalizable dynamic radiance field from monocular videos,

    F. Tian, S. Du, and Y . Duan, “Mononerf: Learning a generalizable dynamic radiance field from monocular videos,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 17 903–17 913

    2023

  44. [52]

    In defense of the eight-point algorithm,

    R. I. Hartley, “In defense of the eight-point algorithm,”IEEE Trans- actions on pattern analysis and machine intelligence, vol. 19, no. 6, pp. 580–593, 1997

    1997

  45. [53]

    Fundamental matrix estimation: A study of error criteria,

    M. E. Fathy, A. S. Hussein, and M. F. Tolba, “Fundamental matrix estimation: A study of error criteria,”Pattern Recognition Letters, vol. 32, no. 2, pp. 383–391, 2011

    2011

  46. [54]

    Hartley and A

    R. Hartley and A. Zisserman,Multiple view geometry in computer vision. Cambridge university press, 2003

    2003

  47. [55]

    Projective manifold gradient layer for deep rotation regression,

    J. Chen, Y . Yin, T. Birdal, B. Chen, L. J. Guibas, and H. Wang, “Projective manifold gradient layer for deep rotation regression,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 6646–6655

    2022

  48. [56]

    Adam: A method for stochastic optimiza- tion,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimiza- tion,”arXiv preprint arXiv:1412.6980, 2014

    2014 arXiv

  49. [57]

    Nerfacc: A general nerf acceleration toolbox,

    R. Li, M. Tancik, and A. Kanazawa, “Nerfacc: A general nerf acceleration toolbox,”arXiv preprint arXiv:2210.04847, 2022

    2022 arXiv

  50. [58]

    Dust3r: Geometric 3d vision made easy,

    S. Wang, V . Leroy, Y . Cabon, B. Chidlovskii, and J. Revaud, “Dust3r: Geometric 3d vision made easy,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 20 697–20 709

    2024

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.