Pith. sign in

REVIEW 4 major objections 6 minor 47 references

MV2: Multi-View Multi-Vehicle Driving Dataset for Novel View Synthesis

T0 review · 4 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read MV2 is a multi-vehicle driving dataset that shows novel-view synthesis quality collapses when train and test views come from different vehicles (car, scooter, drone).

desk verdict Useful new cross-vehicle NVS benchmark, but the headline numbers rest on pose self-consistency rather than absolute pose ground truth; worth a serious referee if the authors commit to validating or softening that claim. read the letter →

arxiv 2608.12442 v1 pith:W54H6F37 submitted 2026-08-12 cs.CV

Sanjay Bhargav Dharavath , Hanvitha Saraswathi Mukkamala , Faizan Farooq Khan , Ioannis Kakogeorgiou , Aditya Arun , C V Jawahar , Zakaria Laskar This is my paper · ORCID
classification cs.CV
keywords novelviewsynthesisdrivingdatasetmulti-vehiclecaptureaerial-to-ground3DGaussianSplattingcameraposeestimationcross-vehiclegeneralizationwide-baselineviews
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

This paper introduces MV2, a driving dataset in which the same urban scenes are captured simultaneously from a car, a scooter, and a drone. The authors use these synchronized multi-vehicle streams to define a novel-view-synthesis benchmark where models are trained on one vehicle's camera and tested on another's, producing much larger viewpoint changes than the same-trajectory splits used by earlier driving benchmarks. They report that every evaluated method loses quality as the train-test camera baseline grows, and that optimization-based rendering approaches clearly outperform feed-forward ones in the wide-baseline regime. The paper argues that such cross-vehicle evaluation measures true extrapolation, which matters for photorealistic driving simulation.

What carries the argument

The load-bearing mechanism is the multi-vehicle capture setup: synchronized car, scooter, and drone cameras looking at the same scene, registered into one coordinate frame by Structure-from-Motion. Poses are certified not by ground truth but by epipolar consistency: human-annotated region correspondences, refined with a dense neural matcher, must keep the mean and maximum epipolar error of every relative pose at or below 30 pixels. That certification step decides which of 200 recorded sequences survive as the 50-scene benchmark, so it is the gatekeeper for every numerical comparison in the paper.

What would settle it

Register a handful of MV2 scenes with surveyed markers or RTK-GPS ground-truth poses, then compare those true poses against the Structure-from-Motion poses that passed the epipolar threshold; if large global pose errors coexist with sub-30-pixel epipolar errors, the certification criterion is insufficient.

Watch

Extended reading notes

Core claim

The central claim is that current novel-view-synthesis models, including dynamic Gaussian-splatting methods that lead on standard splits, degrade sharply when asked to render a view from a different vehicle's trajectory, and that existing single-trajectory benchmarks cannot expose this failure because they only interpolate between nearby frames. On MV2, the best dynamic method drops from 27.01 to 20.23 PSNR when moving from the same-trajectory test split to the car-to-scooter split, and aerial-to-ground splits are harder still. The paper also finds that feed-forward pose estimators produce much larger epipolar errors than optimization-based localization under these wide baselines, so the dataset doubles as a pose-estimation stress test.

Load-bearing premise

The benchmark assumes that a maximum epipolar error of 30 pixels on a few dozen manually annotated correspondences is enough to certify that every camera pose is accurate, because no independent ground-truth pose source exists.

Editorial extensions

If this is right

  • Benchmarks that sample train and test frames from one vehicle trajectory overestimate generalization; MV2-style cross-vehicle splits should be used to measure extrapolation.
  • Aerial-to-ground synthesis, where current methods score below 13 PSNR, is an open problem rather than a solved one.
  • Depth supervision from monocular estimators, when filtered to the 10–30 m range, improves dynamic-scene methods but does not close the wide-baseline gap.
  • Feed-forward pose estimators need to handle large viewpoint gaps before they can replace optimization-based localization in driving scenes.
  • NVS quality on cross-vehicle splits is a direct proxy for whether a simulator can render a plausible lane-change maneuver.

Reading between the lines

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

  • The epipolar-error threshold of 30 pixels could be tested directly: if a subset of scenes were re-certified with surveyed ground-truth poses, the threshold's sufficiency could be quantified rather than assumed.
  • The same benchmark could be extended to closed-loop evaluation, where a rendered scooter or drone view is fed to a perception model; the paper notes downstream uses but does not run that experiment.
  • A natural follow-up is to check whether training feed-forward splatting models on multi-vehicle, wide-baseline data reduces the gap, since the paper only evaluates existing models trained on standard data.
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 / 6 minor

Summary. The paper introduces MV2, a multi-view multi-vehicle driving dataset for novel view synthesis, with synchronized captures from a car, a scooter, and a drone. The authors describe a collection pipeline, SfM-based pose estimation, and a manual correspondence-based epipolar verification step that filters 200 recorded sequences down to 50 scenes. They define two evaluation setups (Eval-Car-Train and Eval-Drone-Train) and report NVS benchmarks for static, dynamic, and feed-forward methods, plus a comparison of COLMAP against feed-forward pose estimators. The headline findings are that NVS performance degrades as the train-test viewpoint disparity increases (e.g., PVG PSNR drops from 27.01 on T_C->C to 20.23 on T_C->S in Table 1) and that feed-forward pose estimators exhibit larger epipolar errors than COLMAP.

Significance. The dataset fills a genuine gap: existing driving NVS benchmarks sample train and test views from one trajectory, while MV2 provides co-registered ground-level and aerial captures of the same dynamic scenes. If the pose verification can be made rigorous, the benchmark would be a valuable resource for studying extrapolation rather than interpolation, and the paper's careful comparison of optimization-based and feed-forward methods, depth-supervision ablations, and dynamic-mask ablations would be useful to the community. The main results are plausible and the release of data and protocol is a strength. However, the paper's central quantitative claims currently rest on pose estimates whose absolute accuracy is explicitly unverified, and the dataset documentation contains inconsistencies that must be fixed before the benchmark can be reproduced.

major comments (4)
  1. [Section 3.5, Eq. (1)] Equation (1) defines the epipolar error as e = |x2^T F x1|, but this is the algebraic epipolar residual, not a pixel distance; the point-to-line distance requires normalization by the length of the epipolar line coefficients. As written, the thresholds em <= 30 used for sequence filtering and the histograms in Fig. 4 are not measured in pixels. Because this threshold is what removes 150 of 200 scenes, the pose-filtering protocol and every downstream conclusion that relies on it need to be re-evaluated with a correctly normalized error, or the implementation should be stated explicitly.
  2. [Section 3.5] The verification of camera poses is self-consistency only. The paper states that 'Since ground-truth poses are unavailable, absolute pose accuracy cannot be directly verified,' and the proposed checks—epipolar consistency on consecutive training edges, one long-range edge, and test-to-nearest-training pairs—cannot rule out global SfM drift or a consistently wrong alignment between the vehicle trajectories. A smoothly drifting reconstruction can satisfy all pairwise epipolar constraints while accumulating large absolute errors, and a mislocalized test image can still satisfy the epipolar constraint with its nearest neighbor. Because every NVS number in Table 1 is computed by rendering at these poses, systematic pose error would directly depress the metrics and could mimic or exaggerate the reported baseline-vs-degradation trend. I recommend validating a subset of scenes against an external reference (e.g., GPS/IMU, surveyed markers, or LiDAR alignment) or, failing that, explicitly reframing the benchmark as a relative/self-consistent evaluation and removing the strong 'pose accuracy' wording.
  3. [Sections 3.1–3.3] The dataset composition is internally inconsistent. Section 3.1 defines only four camera streams (VL, VC, VS, VD), yet Section 3.2 reports '5 camera sequences each' and Section 3.3 mentions a test set T_C->D that was never defined. The abstract's 12,000 images also do not follow from the stated numbers: 50 retained scenes with 5 cameras and 100 frames per camera would yield 25,000 images even before excluding every 5th test frame. The paper must reconcile the number of cameras, define all test sets, and report per-split image counts so the benchmark can be reproduced.
  4. [Section 5.5, Fig. 5] The camera-pose-estimation comparison is evaluated with the same epipolar-error criterion that was used to filter COLMAP poses and select the scenes, so the comparison is not independent: COLMAP is favored by construction, since the 50 scenes were retained precisely because COLMAP's poses passed this check, while feed-forward methods are scored on the same quantity. This does not invalidate the observation that feed-forward methods produce larger epipolar residuals, but the claim that they 'lag behind' optimization-based methods should be presented as a relative statement about this metric, or verified with an external pose reference.
minor comments (6)
  1. [Throughout] Many cross-references are unresolved in the submitted text: 'Section??' appears in Sections 2, 3.3, 4, 5.2, 5.3, 5.4, and 5.5, and 'Fig.??' appears in Section 4; these need to be filled in before publication.
  2. [Section 3.1] Section 3.1 states that the resolution is 1080×1980 while the abstract uses 1080×1920; please correct the inconsistency.
  3. [Section 3.1] Section 3.1 says 'the set of images collected using these three cameras' after defining only two car cameras; the number of cameras per vehicle should be stated unambiguously.
  4. [Section 3.3] In Section 3.3, the sentence beginning 'The following pairs of test sets' is grammatically incomplete and refers to T_C->D, which is not introduced in the Eval-Car-Train setup; please revise.
  5. [Figure 6 caption] Figure 6 caption contains the typo 'BBoth'.
  6. [Section 3.3 / Table 1] The paper does not quantify the actual camera baselines between T_C->C, T_C->L, T_C->S, T_D->C, etc.; since the central trend is attributed to increasing baseline, reporting average baseline distances or another disparity measure would strengthen the claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: MV2's benchmark results are empirical measurements over independent methods, not derivations from their own assumptions.

full rationale

The paper's central claims—NVS performance degrades with increasing train–test viewpoint disparity, aerial-to-ground synthesis is hard, and feed-forward pose estimators underperform optimization-based COLMAP—are empirical measurements over independently implemented methods on a new dataset; none of these results is defined in terms of, or fitted to, the quantity it is said to predict. The test sets TC→C, TC→L, TC→S are ordered by sensor geometry by design, but the observed PSNR/SSIM/LPIPS drops are measured, not implied by that ordering, and the same trend appears across multiple methods including the WOD qualitative cross-lane comparison. No load-bearing self-citation chain appears: the pose-verification procedure (Section 3.5) explicitly states ground-truth poses are unavailable and verifies only epipolar consistency, which is an acknowledged validity limitation of the benchmark rather than a circular derivation; the pose-estimation comparison reuses the same RoMA correspondences as the verification step, but that is a shared evaluation metric, not a fitted parameter renamed as a prediction. The depth-alignment (Eq. 3) and mask ablations are standard preprocessing choices, not predictions. The paper is therefore self-contained against external benchmarks for the purpose of circularity analysis.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The paper is an empirical dataset and benchmark, not a theoretical derivation. The load-bearing assumptions are about pose verification, synchronization, and covisibility; no circular fitted-parameter derivation is present.

free parameters (2)
  • Epipolar error rejection threshold = 30 (maximum epipolar error)
    Hand-chosen threshold used to retain train/test images and filter 200 sequences down to 50 scenes, directly shaping the benchmark's scope and difficulty.
  • Affine depth alignment coefficients = alpha and beta per scene
    Fit per scene in Eq. (3) to align DaV3 monocular depth to COLMAP SfM depth; affects the PVG ablation experiments, not the central benchmark claim.
assumptions (3)
  • domain assumption Epipolar consistency of manually annotated correspondences is a sufficient proxy for absolute pose accuracy.
    Section 3.5 states that ground-truth poses are unavailable and absolute pose accuracy cannot be directly verified, so the entire pose validation chain rests on this proxy.
  • domain assumption Wall-clock synchronization is sufficient to treat car, scooter, and drone captures as observing the same dynamic scene at the same time.
    Section 3.1 and 3.2 describe wall-clock synchronization; dynamic-object epipolar checks are used for validation, but the initial assumption enters the capture protocol itself.
  • domain assumption The 15 nearest training images provide enough covisibility for accurate test image localization.
    Section 3.4 uses this for COLMAP localization and for feeding context views to feed-forward methods; failure of this assumption would bias pose and NVS results.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MV2: Multi-View Multi-Vehicle Driving Dataset for Novel View Synthesis." pith.science (2026). https://pith.science/paper/W54H6F37

@misc{pith2026260812442,
  author       = {Pith},
  title        = {Pith review of: MV2: Multi-View Multi-Vehicle Driving Dataset for Novel View Synthesis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/W54H6F37}},
  note         = {Machine review of arXiv:2608.12442}
}
read the original abstract

Differentiable rendering has advanced novel view synthesis (NVS), yet applying it to real-world driving remains difficult due to sparse capture viewpoints, dynamic objects, and limited multi-trajectory data. We introduce the Multi-View Multi-Vehicle (MV2) dataset and benchmark for evaluating NVS models under large viewpoint changes in dynamic urban scenes. MV2 features synchronized captures from a car, scooter, and drone, each following distinct yet synchronized trajectories. Training NVS methods on one vehicle's camera stream and testing on another enables evaluation under substantially larger viewpoint variations than existing single-trajectory datasets. All sequences are registered via Structure-from-Motion and camera poses verified using manual pixel-level correspondence annotations, yielding 50 high-quality scenes with 12000 images. Benchmarking recent NVS and camera pose estimation methods shows that NVS performance degrades with increasing viewpoint disparity, and that feed-forward pose estimators notably lag behind optimization-based approaches, highlighting MV2 as a rigorous testbed for NVS in driving. The dataset, benchmark protocol, and project resources are available at https://mv2-dataset.github.io/.

Figures

Figures reproduced from arXiv: 2608.12442 by the authors.

Figure 1
Figure 1. Example images and camera trajectories from the proposed MV [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Sample multi-view scenes captured across diverse environments. Each column represents a distinct scene, while each row corresponds to a different acquisition view￾point V C , V D, and V S . The colored borders highlight the viewpoint source of each image, helping visualize cross-view consistency across sensors. consist of small-scale in-bounded scenes captured with a circular camera trajec￾tory around a given object… view at source ↗
Figure 3
Figure 3. Annotation and Pose Evaluation. Representative image pairs from VC -VS (top row) VC -VD (bottom row) with pixel correspondences obtained from annotated region correspondence and RoMA matches and epipolar lines (of randomly sampled RoMA correspondences for visualization purpose). The alignment of correspondences along epipolar lines demonstrates strong geometric consistency and reliable cross-view registration even u… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Epipolar error distribution of mean and maximum epipolar errors ((2)) over 50 scenes. Left: TC→S, Right: TD→C . For the Eval-Drone-Train setup, the training set consists of images cap￾tured from the drone camera V D. The test sets are composed using images from intra-v…
Figure 5
Figure 5. Figure 5: Camera Pose Estimation.COLMAP produces low average epipolar errors (< 30), whereas feed-forward pose estimators yield significantly larger errors (> 50). The number of high-error cases is substantially greater in TD→C due to the wide￾baseline aerial–ground viewpoint ga…
Figure 6
Figure 6. Figure 6: Lane-change scenario rendering comparison. [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]
Figure 7
Figure 7. Figure 7: Qualitative comparison of novel view synthesis across methods. [PITH_FULL_IMAGE:figures/full_fig_p014_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 27 canonical work pages

  1. [1]

    In: CVPR (2022)

    Barron, J.T., Mildenhall, B., Verbin, D., Srinivasan, P.P., Hedman, P.: Mip-nerf 360: Unbounded anti-aliased neural radiance fields. In: CVPR (2022)

    work page 2022

  2. [2]

    arXiv preprint arXiv:2001.10773 (2020)

    Cabon, Y., Murray, N., Humenberger, M.: Virtual kitti 2. arXiv preprint arXiv:2001.10773 (2020)

    arXiv 2020

  3. [3]

    In: CVPR (2020)

    Caesar, H., Bankiti, V., Lang, A.H., Vora, S., Liong, V.E., Xu, Q., Krishnan, A., Pan, Y., Baldan, G., Beijbom, O.: nuscenes: A multimodal dataset for autonomous driving. In: CVPR (2020)

    2020

  4. [4]

    In: CVPR (2024)

    Charatan,D.,Li,S.L.,Tagliasacchi,A.,Sitzmann,V.:pixelsplat:3dgaussiansplats from image pairs for scalable generalizable 3d reconstruction. In: CVPR (2024)

    work page 2024

  5. [5]

    In: ECCV (2024)

    Chen, Y., Xu, H., Zheng, C., Zhuang, B., Pollefeys, M., Geiger, A., Cham, T.J., Cai, J.: Mvsplat: Efficient 3d gaussian splatting from sparse multi-view images. In: ECCV (2024)

    work page 2024

  6. [6]

    arXiv preprint arXiv:2311.18561 (2023)

    Chen, Y., Gu, C., Jiang, J., Zhu, X., Zhang, L.: Periodic vibration gaussian: Dynamic urban scene reconstruction and real-time rendering. arXiv preprint arXiv:2311.18561 (2023)

    arXiv 2023

  7. [7]

    In: CVPR

    Dong, S., Wang, S., Liu, S., Cai, L., Fan, Q., Kannala, J., Yang, Y.: Reloc3r: Large-scale training of relative camera pose regression for generalizable, fast, and accurate visual localization. In: CVPR. pp. 16739–16752 (2025)

    work page 2025

  8. [8]

    In: 3DV (2025)

    Duisterhof, B.P., Zust, L., Weinzaepfel, P., Leroy, V., Cabon, Y., Revaud, J.: MASt3r-sfm: a fully-integrated solution for unconstrained structure-from-motion. In: 3DV (2025)

    work page 2025

Show all 47 references
  1. [9]

    Edstedt, J., Nordström, D., Zhang, Y., Bökman, G., Astermark, J., Larsson, V., Heyden, A., Kahl, F., Wadenbäck, M., Felsberg, M.: Roma v2: Harder better faster denser feature matching (2025),https://arxiv.org/abs/2511.15706

    2025 arXiv

  2. [10]

    In: NeurIPS (2024)

    Fischer, T., Kulhanek, J., Bulo, S.R., Porzi, L., Pollefeys, M., Kontschieder, P.: Dynamic 3d gaussian fields for urban areas. In: NeurIPS (2024)

    2024

  3. [11]

    In: CVPR (2024)

    Fischer, T., Porzi, L., Bulo, S.R., Pollefeys, M., Kontschieder, P.: Multi-level neural scene graphs for dynamic urban environments. In: CVPR (2024)

    2024

  4. [12]

    In: CVPR (2012)

    Geiger, A., Lenz, P., Urtasun, R.: Are we ready for autonomous driving? the kitti vision benchmark suite. In: CVPR (2012)

    2012

  5. [13]

    Cambridge university press (2003)

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

    2003

  6. [14]

    In: NVIDIA Research Whitepapers arXiv:2508.10934 (2025)

    Huang, J., Zhou, Q., Rabeti, H., Korovko, A., Ling, H., Ren, X., Shen, T., Gao, J., Slepichev, D., Lin, C.H., Ren, J., Xie, K., Biswas, J., Leal-Taixe, L., Fidler, S.: Vipe: Video pose engine for 3d geometric perception. In: NVIDIA Research Whitepapers arXiv:2508.10934 (2025)

    2025 arXiv

  7. [15]

    arXiv preprint arXiv:2405.20323 (2024)

    Huang, N., Wei, X., Zheng, W., An, P., Lu, M., Zhan, W., Tomizuka, M., Keutzer, K., Zhang, S.:S3Gaussian: Self-supervised street gaussians for autonomous driving. arXiv preprint arXiv:2405.20323 (2024)

    2024 arXiv

  8. [16]

    arXiv preprint arXiv:2509.13414 (2025) MV2: Multi-Vehicle Driving Dataset 17

    Keetha, N., Müller, N., Schönberger, J., Porzi, L., Zhang, Y., Fischer, T., Knapitsch, A., Zauss, D., Weber, E., Antunes, N., et al.: MapAnything: Universal feed-forward metric 3d reconstruction. arXiv preprint arXiv:2509.13414 (2025) MV2: Multi-Vehicle Driving Dataset 17

    2025 arXiv

  9. [17]

    TOG (2023)

    Kerbl, B., Kopanas, G., Leimkühler, T., Drettakis, G.: 3d gaussian splatting for real-time radiance field rendering. TOG (2023)

    2023

  10. [18]

    In: ICCV (2023)

    Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., Xiao, T., Whitehead, S., Berg, A.C., Lo, W.Y., et al.: Segment anything. In: ICCV (2023)

    2023

  11. [19]

    In: arXiv preprint arXiv:2406.18360 (2024)

    Li, H., Yuan, M., Zhang, Y., Wu, C., Zhao, C., Song, C., Feng, H., Ding, E., Zhang, D., Wang, J.: Xld: a cross-lane dataset for benchmarking novel driving view synthesis. In: arXiv preprint arXiv:2406.18360 (2024)

    2024 arXiv

  12. [20]

    In: arXiv preprint arXiv:2503.12552 (2025)

    Li, T., Qiu, Y., Wu, Z., Lindström, C., Su, P., Nießner, M., Li, H.: Mtgs: Multi- traversal gaussian splatting. In: arXiv preprint arXiv:2503.12552 (2025)

    2025 arXiv

  13. [21]

    org/abs/2511.10647

    Lin, H., Chen, S., Liew, J., Chen, D.Y., Li, Z., Shi, G., Feng, J., Kang, B.: Depth anything 3: Recovering the visual space from any views (2025),https://arxiv. org/abs/2511.10647

    2025 arXiv

  14. [22]

    In: CVPR (2024)

    Ling, L., Sheng, Y., Tu, Z., Zhao, W., Xin, C., Wan, K., Yu, L., Guo, Q., Yu, Z., Lu, Y., et al.: Dl3dv-10k: A large-scale scene dataset for deep learning-based 3d vision. In: CVPR (2024)

    2024

  15. [23]

    In: CVPR (2025)

    Liu, Y., Fan, K., Yu, W., Li, C., Lu, H., Yuan, Y.: Monosplat: Generalizable 3d gaussian splatting from monocular depth foundation models. In: CVPR (2025)

    2025

  16. [24]

    In: ECCV (2024)

    Ljungbergh, W., Tonderski, A., Johnander, J., Caesar, H., Åström, K., Felsberg, M., Petersson, C.: Neuroncap: Photorealistic closed-loop safety testing for au- tonomous driving. In: ECCV (2024)

    2024

  17. [25]

    In: AAAI (2025)

    Ma, X., Zhang, J., Lu, P., Xu, S., Pan, C.: Novel view synthesis under large- deviation viewpoint for autonomous driving. In: AAAI (2025)

    2025

  18. [26]

    Commu- nications of the ACM (2021)

    Mildenhall, B., Srinivasan, P.P., Tancik, M., Barron, J.T., Ramamoorthi, R., Ng, R.: Nerf: Representing scenes as neural radiance fields for view synthesis. Commu- nications of the ACM (2021)

    2021

  19. [27]

    arXiv preprint arXiv:2502.15635 (2025)

    Ni, Z., Du, S., Hou, Z., Wu, C., Yang, S.: Para-lane: Multi-lane dataset reg- istering parallel scans for benchmarking novel view synthesis. arXiv preprint arXiv:2502.15635 (2025)

    2025 arXiv

  20. [28]

    In: 3DV (2025)

    Ni, Z., Du, S., Hou, Z., Wu, C., Yang, S.: Para-lane: Multi-lane dataset registering parallel scans for benchmarking novel view synthesis. In: 3DV (2025)

    2025

  21. [29]

    In: CVPR (2025)

    Peng, C., Zhang, C., Wang, Y., Xu, C., Xie, Y., Zheng, W., Keutzer, K., Tomizuka, M., Zhan, W.: Desire-gs: 4d street gaussians for static-dynamic decomposition and surface reconstruction for urban driving scenes. In: CVPR (2025)

    2025

  22. [30]

    In: CVPR (2016)

    Schönberger, J.L., Frahm, J.M.: Structure-from-motion revisited. In: CVPR (2016)

    2016

  23. [31]

    In: CVPR (2020)

    Sun, P., Kretzschmar, H., Dotiwalla, X., Chouard, A., Patnaik, V., Tsui, P., Guo, J., Zhou, Y., Chai, Y., Caine, B., et al.: Scalability in perception for autonomous driving: Waymo open dataset. In: CVPR (2020)

    2020

  24. [32]

    In: CVPR (2025)

    Tang, J., Gao, Y., Yang, D., Yan, L., Yue, Y., Yang, Y.: Dronesplat: 3d gaussian splatting for robust 3d reconstruction from in-the-wild drone imagery. In: CVPR (2025)

    2025

  25. [33]

    In: CVPR (2024)

    Tonderski, A., Lindström, C., Hess, G., Ljungbergh, W., Svensson, L., Petersson, C.: Neurad: Neural rendering for autonomous driving. In: CVPR (2024)

    2024

  26. [34]

    In: CVPR (2023)

    Turki, H., Zhang, J.Y., Ferroni, F., Ramanan, D.: Suds: Scalable urban dynamic scenes. In: CVPR (2023)

    2023

  27. [35]

    In: CVPR (2025)

    Vuong, K., Ghosh, A., Ramanan, D., Narasimhan, S., Tulsiani, S.: Aerialmegadepth: Learning aerial-ground reconstruction and view synthesis. In: CVPR (2025)

    2025

  28. [36]

    In: CVPR (2025) 18 S

    Wang, J., Chen, M., Karaev, N., Vedaldi, A., Rupprecht, C., Novotny, D.: Vggt: Visual geometry grounded transformer. In: CVPR (2025) 18 S. B. Dharavath et al

    2025

  29. [37]

    In: NeurIPS (2021)

    Wilson, B., Qi, W., Agarwal, T., Lambert, J., Singh, J., Khandelwal, S., Pan, B., Kumar, R., Hartnett, A., Pontes, J.K., Ramanan, D., Carr, P., Hays, J.: Argoverse 2: Next generation datasets for self-driving perception and forecasting. In: NeurIPS (2021)

    2021

  30. [38]

    In: CVPR (2025)

    Wu, W., He, H., Zhang, C., He, J., Zhao, S.Z., Gong, R., Li, Q., Zhou, B.: Towards autonomous micromobility through scalable urban simulation. In: CVPR (2025)

    2025

  31. [39]

    In: CVPR (2025)

    Xie, Z., Liu, Z., Peng, Z., Wu, W., Zhou, B.: Vid2sim: Realistic and interactive simulation from video for urban navigation. In: CVPR (2025)

    2025

  32. [40]

    In: CVPR (2024)

    Xu, H., Peng, S., Wang, F., Blum, H., Barath, D., Geiger, A., Pollefeys, M.: Depth- splat: Connecting gaussian splatting and depth. In: CVPR (2024)

    2024

  33. [41]

    In: CVPR (2022)

    Xu, Q., Xu, Z., Philip, J., Bi, S., Shu, Z., Sunkavalli, K., Neumann, U.: Point-nerf: Point-based neural radiance fields. In: CVPR (2022)

    2022

  34. [42]

    In: ECCV (2024)

    Yan, Y., Lin, H., Zhou, C., Wang, W., Sun, H., Zhan, K., Lang, X., Zhou, X., Peng, S.: Street gaussians: Modeling dynamic urban scenes with gaussian splatting. In: ECCV (2024)

    2024

  35. [43]

    In: ICLR (2025)

    Yang, J., Huang, J., Ivanovic, B., Chen, Y., Wang, Y., Li, B., You, Y., Sharma, A., Igl, M., Karkus, P., Xu, D., Wang, Y., Pavone, M.: STORM: Spatio-temporal reconstruction model for large-scale outdoor scenes. In: ICLR (2025)

    2025

  36. [44]

    In: ICLR (2024)

    Yang, J., Ivanovic, B., Litany, O., Weng, X., Kim, S.W., Li, B., Che, T., Xu, D., Fidler, S., Pavone, M., Wang, Y.: EmerneRF: Emergent spatial-temporal scene decomposition via self-supervision. In: ICLR (2024)

    2024

  37. [45]

    In: CVPR (2018)

    Zhang, R., Isola, P., Efros, A.A., Shechtman, E., Wang, O.: The unreasonable effectiveness of deep features as a perceptual metric. In: CVPR (2018)

    2018

  38. [46]

    TOG (2018)

    Zhou, T., Tucker, R., Flynn, J., Fyffe, G., Snavely, N.: Stereo magnification: Learn- ing view synthesis using multiplane images. TOG (2018)

    2018

  39. [47]

    In: CVPR (2024)

    Zhou, X., Lin, Z., Shan, X., Wang, Y., Sun, D., Yang, M.H.: Drivinggaussian: Composite gaussian splatting for surrounding dynamic autonomous driving scenes. In: CVPR (2024)

    2024

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.