Pith. sign in

REVIEW 3 major objections 6 minor 32 references

LiDAR-EDIT: LiDAR Data Generation by Editing the Object Layouts in Real-World Scenes

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Editing real LiDAR scans produces controllable synthetic data that improves detector training.

desk verdict A genuinely novel LiDAR editing paradigm with a clean spherical-voxelization core, but the inpainting evidence skips the real de-occlusion masks that matter most. read the letter →

arxiv 2412.00592 v3 pith:OU3347YK submitted 2024-11-30 cs.CV cs.AIcs.LGcs.RO

Shing-Hei Ho , Bao Thach , Minghan Zhu This is my paper · ORCID
classification cs.CVcs.AIcs.LGcs.RO
keywords LiDARsceneeditingsyntheticgenerationbackgroundinpaintingsphericalvoxelizationobjectinsertionautonomousdrivingpointcloudcompletiondiscretelatenttransformer
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

LiDAR-EDIT proposes to generate synthetic LiDAR data not by rendering from scratch but by editing real scans: it removes vehicles, fills the occluded background with a generative inpainting model, and inserts point-cloud objects from a library at user-chosen poses. The paper argues this preserves real-world background realism, gives full control over object count, type, and pose, and produces object labels for free, avoiding the domain gap of simulators and the limited controllability of end-to-end generation. The load-bearing mechanism is spherical voxelization, which discretizes space by range, azimuth, and elevation so occlusion and scan density follow real LiDAR projective geometry by construction. Experiments on a standard driving dataset show that detectors pretrained on edited scenes and then fine-tuned on real data outperform detectors trained on real data alone, with a small measured domain gap.

What carries the argument

Spherical voxelization, which discretizes space around the sensor into $(r,\theta,\phi)$ bins so that each LiDAR ray is a column of voxels, is the core object; it makes occlusion and beam-pattern consistency explicit, so removal masks and insertion resampling are computed by simple ray-column operations. Around it sit a vector-quantized variational autoencoder and a masked bidirectional transformer trained on azimuth-radius bird's-eye-view latent maps for background inpainting, an object library built with a pretrained point-cloud completion network, and an insertion routine that aligns objects to the ground and removes occluded background voxels.

What would settle it

Apply the full removal-and-inpainting pipeline to scans where the area behind a real vehicle is later observed from a nearby second LiDAR pass, then compare the inpainted points with the actually observed points: a large discrepancy in geometry or density would falsify the realism claim for object removal.

Watch

Extended reading notes

Core claim

The paper's central claim is that realistic, usable LiDAR point clouds can be manufactured by surgical editing of real scans rather than by full-scene synthesis. Object removal is treated as de-occlusion masking plus generative background inpainting; object insertion is completed-shape placement with spherical-voxel resampling and occlusion resolution. The result is a synthetic scan with the original background, a newly specified object layout, and labels inherited from the object library. The authors support the claim with detector-based evaluation: the same detector's mean average precision is similar on real and edited validation sets, and pretraining on edited data improves downstream detection over real-only training.

Load-bearing premise

The inpainting network is trained on masks over random object-free voxels and is assumed to generalize to the larger, structurally different de-occlusion masks left by actual objects; if it does not, every edited scene inherits unrealistic background artifacts.

Editorial extensions

If this is right

  • Users can specify how many objects, of which types, and at which poses appear in a scan while keeping the original real environment.
  • Generated scans come with object labels, which end-to-end LiDAR generators generally do not provide.
  • Pretraining an object detector on edited data before fine-tuning on real data improves mean average precision over training on real data alone.
  • The method can create counterfactual scenes with object layouts that differ greatly from the recorded scene, enabling controlled comparisons.
  • Because only modified regions are synthesized, the approach scales to existing large datasets without physics engines or full-scene reconstruction.

Reading between the lines

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

  • If the realism claim is right, a direct stress test would be to evaluate background inpainting on real object de-occlusion masks instead of the average-size synthetic masks used in the paper; this is the point where the training and inference distributions differ most.
  • The same editing pipeline could be pointed at safety-critical counterfactual queries, such as moving a pedestrian into the ego lane, once the object library and completion models support non-car categories reliably.
  • Modeling secondary returns and ray dropping, which the paper lists as future work, should close the remaining gap between inserted cars and real cars and likely raise the car-category average precision further.
  • Because object labels are known by construction, edited data could serve as free annotated training and testing data for detection, tracking, and motion forecasting without manual labeling.
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

3 major / 6 minor

Summary. The paper proposes LiDAR-EDIT, a pipeline that edits real LiDAR scans by removing objects, inpainting the occluded background with a MaskGIT-style generative model, completing partial object scans into a library, and re-inserting objects at user-specified poses using spherical voxelization to enforce LiDAR projective geometry. The method is evaluated on nuScenes in three ways: background inpainting quality on synthetic masks, object removal effectiveness via a pretrained detector, and object-insertion realism via domain-gap metrics and a downstream detection pretraining experiment. The paper claims small domain gap and practical value for autonomous-driving tasks.

Significance. The paradigm itself is a useful and reasonably novel contribution: editing real scans rather than generating from scratch preserves background realism and offers controllable counterfactual object layouts. The spherical-voxelization formulation is clean and gives occlusion handling and LiDAR-consistent resampling by construction, which is a genuine strength. The paper also provides open-sourced code and uses standard perceptual/statistical metrics. However, the evidence supporting the central realism claim is incomplete: the inpainting model is trained and evaluated on masks that differ structurally from the de-occlusion masks used at inference, and the downstream experiment confounds extra training epochs with synthetic-data value. With additional targeted evaluation the contribution could be solid, but as it stands the claims are not fully supported.

major comments (3)
  1. [V-A] The background inpainting evaluation uses only artificially created masks: a nominal bounding box of average size at 10 m placed in object-free sectors (Sec V-A). Meanwhile, training uses masks on "random object-free voxels" (Sec IV-C2) and inference uses de-occlusion masks generated from real objects (Sec IV-C1), which are larger, more contiguous, and have radial occlusion patterns behind the object. The paper provides no evaluation of inpainting fidelity on these real de-occlusion masks; the object-removal experiment in Tab. II only counts residual detected objects and does not measure whether the inpainted background is realistic. Because the realism of every edited scene depends on background inpainting after real object removal, this train/test mask mismatch is load-bearing and leaves the primary realism claim unverified.
  2. [V-B.2] The downstream detection comparison confounds the effect of synthetic pretraining with the effect of additional training epochs: the baseline model is trained on real nuScenes for 40 epochs, while the pretrained model is trained on synthetic data for 15 epochs and then on real data for 40 more epochs, for a total of 55 epochs. Without a control that trains on real data for 55 epochs (or otherwise matches total compute), the observed mAP improvement from 0.361 to 0.368 cannot be attributed to the synthetic data. This is a central claim of the paper, so the experiment needs to be re-run with matched training budgets.
  3. [Tables I-III] All quantitative results are reported as single numbers without error bars, multiple seeds, or statistical significance tests. The key claims of a small domain gap (mAP 0.361 vs. 0.340) and downstream improvement (mAP 0.361 vs. 0.368) involve small differences that could easily arise from training stochasticity. The paper should report mean and variance over at least three seeds, especially for the detector-based evaluations in Tabs. II and III.
minor comments (6)
  1. [V-A] Typo: "background poins" should be "background points".
  2. [V-A] The inpainting baseline (copy-tile of a neighboring object-free sector) is very weak; while the task has no established prior method, a stronger baseline (e.g., a range-image diffusion model or UltraLidar-style generation restricted to the mask) would make the quantitative improvements more convincing.
  3. [IV-C2] The paper does not report the distribution of training masks (sizes, shapes, contiguity, distances) used for background inpainting, which makes it difficult to assess how different the inference-time de-occlusion masks are. A figure or table summarizing training mask statistics would help.
  4. [IV-D] The object library relies on AnchorFormer, but the paper does not state what dataset this completion network was pretrained on or how its domain gap to nuScenes objects is handled. This should be clarified.
  5. [V-B.1] The higher car AP on synthetic data (0.785 vs. 0.672) is acknowledged as possibly due to simpler, ray-drop-free inserted objects, but the paper should quantify this (e.g., by comparing insertion point density or detection confidence distributions) rather than leaving it as a speculation.
  6. [III] The problem formulation assumes every object in the target set appears somewhere in the training data; the paper should discuss the coverage of the object library and possible failure modes for rare or unseen categories.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the pipeline is a constructive editing system evaluated empirically; no prediction reduces to fitted inputs or to self-citations.

full rationale

The paper's central mechanism, spherical voxelization, is a constructive discretization that mirrors LiDAR ray geometry; calling it 'correct by construction' is a description of the construction, not a fitted prediction. The generative modules (background inpainting, object completion, detection) are all based on externally published methods (UltraLidar, MaskGIT, AnchorFormer, VoxelNext), not on self-citations, and the downstream detection experiment is an empirical comparison rather than a quantity forced by fitting. The inpainting evaluation uses artificial masks of average object size at 10 m rather than real de-occlusion masks, and the perceptual metrics use feature extractors pretrained on nuScenes, the same dataset used to train the inpainting model; these are legitimate generalizability and evaluation-domain concerns, but they do not make any derivation equivalent to its inputs by construction. No equation or fitted parameter is renamed as a prediction, and no load-bearing claim reduces to a self-citation chain. The honest finding is therefore no significant circularity.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical entities; it is a pipeline of learned modules and geometric heuristics. The main hand-chosen parameters are the spherical voxelization geometry, the evaluation mask design, and the pretraining schedule. The load-bearing assumptions concern scene decomposition, object library coverage, first-return-only LiDAR modeling, flat-ground placement, and the train-to-test transfer of inpainting masks.

free parameters (4)
  • Spherical voxelization resolution and field ranges = 512x512x32; 50 m radius, 360 deg azimuth, 79.3 to 121 deg elevation
    Hand-chosen discretization that defines the 'correct by construction' occlusion and resampling behavior; the fidelity of the result depends on these values matching the HDL32E sensor pattern (Sec IV-B, V).
  • Inpainting evaluation mask geometry = Average vehicle bounding box at 10 m distance
    The quantitative inpainting evaluation is performed only on masks created from an average-sized box rotated into object-free sectors, so reported metrics may not reflect real object de-occlusion (Sec V-A).
  • Pose perturbation range for pretraining data = Up to 2.5 m translation and 45 deg yaw
    Synthetic training data perturbs insertion poses by these bounds; the measured downstream value could depend on this range (Sec V-B.2).
  • Pretraining schedule for downstream experiment = 15 epochs synthetic, then 40 epochs real
    The claimed practical value is demonstrated for this specific schedule; no ablation shows sensitivity to the ratio (Sec V-B.2).
assumptions (5)
  • domain assumption The scene decomposes into static background S and foreground dynamic objects D, with segmentation masks separating them.
    Used in Sec III and throughout the removal stage (Sec IV-C); requires that vehicles can be cleanly separated and that the background is static and coherent once objects are removed.
  • domain assumption All objects that may be inserted must already appear in the training data's object library.
    Stated in Sec III: 'We assume that all possible objects in DT are observed in some Pi.' Limits the method to object types and shapes seen in the source dataset.
  • domain assumption LiDAR can be modeled as a first-return-only spherical ray pattern, ignoring ray-drop, beam divergence, and secondary returns.
    The spherical voxelization occlusion rule (Sec IV-B, IV-E2) keeps only the nearest occupied voxel per ray; the paper lists ray-drop as future work in Sec VI.
  • domain assumption Inserted objects have zero pitch and roll and are aligned to ground via the nearest ground point.
    Sec IV-E1 assumes flat ground contact, which is typical of annotated driving datasets but not guaranteed for arbitrary roads.
  • ad hoc to paper The inpainting model trained on random object-free masks transfers to real de-occlusion masks of actual objects.
    Training masks are random object-free voxels (Sec IV-C2), while inference uses de-occlusion masks from real objects; the evaluation uses artificial masks (Sec V-A) and never validates this transfer directly.

how reviews work

0 comments
Cite this review

Pith. "Pith review of LiDAR-EDIT: LiDAR Data Generation by Editing the Object Layouts in Real-World Scenes." pith.science (2026). https://pith.science/paper/OU3347YK

@misc{pith2026241200592,
  author       = {Pith},
  title        = {Pith review of: LiDAR-EDIT: LiDAR Data Generation by Editing the Object Layouts in Real-World Scenes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OU3347YK}},
  note         = {Machine review of arXiv:2412.00592}
}
read the original abstract

We present LiDAR-EDIT, a novel paradigm for generating synthetic LiDAR data for autonomous driving. Our framework edits real-world LiDAR scans by introducing new object layouts while preserving the realism of the background environment. Compared to end-to-end frameworks that generate LiDAR point clouds from scratch, LiDAR-EDIT offers users full control over the object layout, including the number, type, and pose of objects, while keeping most of the original real-world background. Our method also provides object labels for the generated data. Compared to novel view synthesis techniques, our framework allows for the creation of counterfactual scenarios with object layouts significantly different from the original real-world scene. LiDAR-EDIT uses spherical voxelization to enforce correct LiDAR projective geometry in the generated point clouds by construction. During object removal and insertion, generative models are employed to fill the unseen background and object parts that were occluded in the original real LiDAR scans. Experimental results demonstrate that our framework produces realistic LiDAR scans with practical value for downstream tasks.

Figures

Figures reproduced from arXiv: 2412.00592 by the authors.

Figure 1
Figure 1. Overview of our novel LiDAR editing paradigm. Given the point cloud of a real-world LiDAR scan, we want to freely change the objects and their poses while preserving the background environment. This requires filling the background when objects are removed, and handling occlusion and LiDAR scan projection when new objects are inserted. Edited points are highlighted in red. B. LiDAR novel-view synthesis LiDAR novel-vi… view at source ↗
Figure 2
Figure 2. Overview of our novel LiDAR-EDIT framework for LiDAR editing. Asterisk signs denote modules with generative models [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. (Left) Spherical voxelization discretizes the space based on radius r (distance from the origin), azimuth θ (horizontal angle), and elevation ϕ (vertical angle). (Right) Occlusion handling in spherical representation is straightforward. If a voxel at coordinate (r, θ, ϕ) is occupied (Green), all voxels with the same azimuth and elevation but a larger radius ((r ′ , θ, ϕ) where r ′ > r) will be occluded (Red). (Sec. … view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Overview of the background inpainting model. There are two stages in the training. (a) Use a VQ-VAE model to learn a discrete latent map in the bird’s eye view. The colors represent the discrete latent codes. (b) Learn a multi-step autoregressive generation model that …
Figure 5
Figure 5. Figure 5: Illustration of the inpainting mask creation process in the background inpainting experiment. (a) shows the object-free sectors in the original scan. (b) shows that we use a nominal bounding box of the average size at 10 meters distance to create the mask. The bounding…
Figure 6
Figure 6. Figure 6: Qualitative results of background inpainting on nuScenes dataset. Red highlights the points in the inpainting mask. Our prediction fits naturally with the surrounding environment and is close to the actual scans. Data generation method Ours UltraLidar [19] Number of ob…
Figure 7
Figure 7. Figure 7: Object insertion qualitative results. Inserted points are highlighted in red. Training data Testing data mAP ↑ car ↑ non-car ↑ nuScenes Ours 0.340 0.785 0.290 nuScenes nuScenes 0.361 0.672 0.326 Ours + nuScenes nuScenes 0.368 0.672 0.334 TABLE III: Object detection per…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 26 canonical work pages

  1. [1]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” Advances in neural information processing systems , vol. 33, pp. 6840–6851, 2020. 1

    work page 2020

  2. [2]

    Unsupervised representation learning with deep convolutional generative adversarial networks,

    A. Radford, L. Metz, and S. Chintala, “Unsupervised representation learning with deep convolutional generative adversarial networks,” arXiv preprint arXiv:1511.06434 , 2015. 1

    arXiv 2015

  3. [3]

    Deep generative modeling of lidar data,

    L. Caccia, H. Van Hoof, A. Courville, and J. Pineau, “Deep generative modeling of lidar data,” in 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2019, pp. 5034–

    work page 2019

  4. [4]

    Learning to generate realistic lidar point clouds,

    V . Zyrianov, X. Zhu, and S. Wang, “Learning to generate realistic lidar point clouds,” in European Conference on Computer Vision. Springer, 2022, pp. 17–35. 1, 2

    work page 2022

  5. [5]

    Neural lidar fields for novel view synthesis,

    S. Huang, Z. Gojcic, Z. Wang, F. Williams, Y . Kasten, S. Fidler, K. Schindler, and O. Litany, “Neural lidar fields for novel view synthesis,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 18 236–18 246. 1, 2

    work page 2023

  6. [6]

    A lidar point cloud generator: from a virtual world to autonomous driving,

    X. Yue, B. Wu, S. A. Seshia, K. Keutzer, and A. L. Sangiovanni- Vincentelli, “A lidar point cloud generator: from a virtual world to autonomous driving,” in Proceedings of the 2018 ACM on interna- tional conference on multimedia retrieval , 2018, pp. 458–464. 1

    work page 2018

  7. [7]

    Unreal engine

    Epic Games, “Unreal engine.” [Online]. Available: https://www. unrealengine.com 1

    work page

  8. [8]

    Unity: A general platform for intelligent agents,

    A. Juliani, “Unity: A general platform for intelligent agents,” arXiv preprint arXiv:1809.02627, 2018. 1

    arXiv 2018

Show all 32 references
  1. [9]

    Carla: An open urban driving simulator,

    A. Dosovitskiy, G. Ros, F. Codevilla, A. Lopez, and V . Koltun, “Carla: An open urban driving simulator,” in Conference on robot learning . PMLR, 2017, pp. 1–16. 1

    2017

  2. [10]

    Lidarsim: Realistic lidar simulation by leveraging the real world,

    S. Manivasagam, S. Wang, K. Wong, W. Zeng, M. Sazanovich, S. Tan, B. Yang, W.-C. Ma, and R. Urtasun, “Lidarsim: Realistic lidar simulation by leveraging the real world,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020, pp. 11 167–11 176. 1

    2020

  3. [11]

    Towards zero domain gap: A comprehensive study of realistic lidar simulation for autonomy testing,

    S. Manivasagam, I. A. B ˆarsan, J. Wang, Z. Yang, and R. Urtasun, “Towards zero domain gap: A comprehensive study of realistic lidar simulation for autonomy testing,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 8272–8282. 1

    2023

  4. [12]

    Lidar-nerf: Novel lidar view synthesis via neural radiance fields,

    T. Tao, L. Gao, G. Wang, Y . Lao, P. Chen, H. Zhao, D. Hao, X. Liang, M. Salzmann, and K. Yu, “Lidar-nerf: Novel lidar view synthesis via neural radiance fields,” arXiv preprint arXiv:2304.10406 , 2023. 2

    2023 arXiv

  5. [13]

    Unisim: A neural closed-loop sensor simulator,

    Z. Yang, Y . Chen, J. Wang, S. Manivasagam, W.-C. Ma, A. J. Yang, and R. Urtasun, “Unisim: A neural closed-loop sensor simulator,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 1389–1399. 2

    2023

  6. [14]

    Nerf-lidar: Generating realistic lidar point clouds with neural radiance fields,

    J. Zhang, F. Zhang, S. Kuang, and L. Zhang, “Nerf-lidar: Generating realistic lidar point clouds with neural radiance fields,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 38, no. 7, 2024, pp. 7178–7186. 2

    2024

  7. [15]

    Lidar4d: Dynamic neural fields for novel space-time view lidar synthesis,

    Z. Zheng, F. Lu, W. Xue, G. Chen, and C. Jiang, “Lidar4d: Dynamic neural fields for novel space-time view lidar synthesis,” in Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 5145–5154. 2

    2024

  8. [16]

    Dynamic lidar re-simulation using compositional neural fields,

    H. Wu, X. Zuo, S. Leutenegger, O. Litany, K. Schindler, and S. Huang, “Dynamic lidar re-simulation using compositional neural fields,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 19 988–19 998. 2

    2024

  9. [17]

    Alignmif: Geometry-aligned multimodal implicit field for lidar-camera joint synthesis,

    T. Tao, G. Wang, Y . Lao, P. Chen, J. Liu, L. Lin, K. Yu, and X. Liang, “Alignmif: Geometry-aligned multimodal implicit field for lidar-camera joint synthesis,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 21 230–21 240. 2

    2024

  10. [18]

    Towards realistic scene gener- ation with lidar diffusion models,

    H. Ran, V . Guizilini, and Y . Wang, “Towards realistic scene gener- ation with lidar diffusion models,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 14 738–14 748. 2, 5

    2024

  11. [19]

    Learning compact representations for lidar completion and generation,

    Y . Xiong, W.-C. Ma, J. Wang, and R. Urtasun, “Learning compact representations for lidar completion and generation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recog- nition, 2023, pp. 1074–1083. 2, 3, 5

    2023

  12. [20]

    Maskgit: Masked generative image transformer,

    H. Chang, H. Zhang, L. Jiang, C. Liu, and W. T. Freeman, “Maskgit: Masked generative image transformer,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion, 2022, pp. 11 315–11 325. 2, 3

    2022

  13. [21]

    Learning unsupervised world models for autonomous driving via discrete diffusion,

    L. Zhang, Y . Xiong, Z. Yang, S. Casas, R. Hu, and R. Urtasun, “Learning unsupervised world models for autonomous driving via discrete diffusion,” arXiv preprint arXiv:2311.01017 , 2023. 2

    2023 arXiv

  14. [22]

    Lidardm: Generative lidar simulation in a generated world,

    V . Zyrianov, H. Che, Z. Liu, and S. Wang, “Lidardm: Generative lidar simulation in a generated world,” arXiv preprint arXiv:2404.02903 ,

  15. [23]

    Tulip: Transformer for upsampling of lidar point clouds,

    B. Yang, P. Pfreundschuh, R. Siegwart, M. Hutter, P. Moghadam, and V . Patil, “Tulip: Transformer for upsampling of lidar point clouds,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 15 354–15 364. 2

    2024

  16. [24]

    Scaling diffusion models to real-world 3d lidar scene completion,

    L. Nunes, R. Marcuzzi, B. Mersch, J. Behley, and C. Stachniss, “Scaling diffusion models to real-world 3d lidar scene completion,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 14 770–14 780. 2

    2024

  17. [25]

    Genmm: Geometrically and temporally consistent multimodal data generation for video and lidar,

    B. Singh, V . Kulharia, L. Yang, A. Ravichandran, A. Tyagi, and A. Shrivastava, “Genmm: Geometrically and temporally consistent multimodal data generation for video and lidar,” arXiv preprint arXiv:2406.10722, 2024. 2

    2024 arXiv

  18. [26]

    Neural discrete representation learning,

    A. Van Den Oord, O. Vinyals, et al., “Neural discrete representation learning,” Advances in neural information processing systems, vol. 30,

  19. [27]

    Anchorformer: Point cloud completion from discriminative nodes,

    Z. Chen, F. Long, Z. Qiu, T. Yao, W. Zhou, J. Luo, and T. Mei, “Anchorformer: Point cloud completion from discriminative nodes,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 13 581–13 590. 4, 5

    2023

  20. [28]

    nuscenes: A multimodal dataset for autonomous driving,

    H. Caesar, V . Bankiti, A. H. Lang, S. V ora, V . E. Liong, Q. Xu, A. Krishnan, Y . Pan, G. Baldan, and O. Beijbom, “nuscenes: A multimodal dataset for autonomous driving,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2020, pp. 11 621–...

    2020

  21. [29]

    Scalability in perception for autonomous driving: Waymo open dataset,

    P. Sun, H. Kretzschmar, X. Dotiwalla, A. Chouard, V . Patnaik, P. Tsui, J. Guo, Y . Zhou, Y . Chai, B. Caine, et al., “Scalability in perception for autonomous driving: Waymo open dataset,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , ...

    2020

  22. [30]

    4d spatio-temporal convnets: Minkowski convolutional neural networks,

    C. Choy, J. Gwak, and S. Savarese, “4d spatio-temporal convnets: Minkowski convolutional neural networks,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2019, pp. 3075–3084. 5

    2019

  23. [31]

    Search- ing efficient 3d architectures with sparse point-voxel convolution,

    H. Tang, Z. Liu, S. Zhao, Y . Lin, J. Lin, H. Wang, and S. Han, “Search- ing efficient 3d architectures with sparse point-voxel convolution,” in European conference on computer vision . Springer, 2020, pp. 685–

    2020

  24. [32]

    V oxelnext: Fully sparse voxelnet for 3d object detection and tracking,

    Y . Chen, J. Liu, X. Zhang, X. Qi, and J. Jia, “V oxelnext: Fully sparse voxelnet for 3d object detection and tracking,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 21 674–21 683. 5, 6

    2023

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.