Pith. sign in

REVIEW 5 major objections 6 minor 70 references

LaGen: Towards Autoregressive LiDAR Scene Generation

T0 review · 5 major / 6 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read LaGen autoregressively generates long-horizon LiDAR driving scenes frame by frame from a single starting frame.

desk verdict LaGen is a genuinely new interactive LiDAR-world-model framework, but the quantitative support for its headline claims is undercut by an impossible MMD and an oracle-conditioning problem in the baselines. read the letter →

arxiv 2511.21256 v2 pith:23G7G2E3 submitted 2025-11-26 cs.CV

Sizhuo Zhou , Xiaosong Jia , Fanrui Zhang , Junjie Li , Juyong Zhang , Yukang Feng , Jianwen Sun , Songbur Wong
show 2 more authors
Junqi You Junchi Yan
This is my paper
classification cs.CV
keywords LiDARscenegenerationautoregressivelatentdiffusionmodelrangeimageworldautonomousdrivingnuSceneslong-horizonsimulation
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

LaGen claims to be the first framework that can generate long-horizon LiDAR scenes autoregressively, one frame at a time, starting from a single LiDAR frame and using bounding-box and ego-state conditions. Prior LiDAR generators only create single frames, while predictors need multiple historical frames and lack interactivity. LaGen shows lower Chamfer distance and ray-depth errors than prediction baselines, especially at later timestamps, and supports interactive edits like moving or removing objects. This matters because it opens a path toward closed-loop simulation and data augmentation for autonomous driving.

What carries the argument

The central object is a latent diffusion model applied to corrected range images, where multiple conditions—previous frame latent features adjusted by relative transform, object-level bounding-box masks injected via cross-attention, and ego-state embeddings—guide denoising. The SDE module transforms previous-frame points into estimated current-frame foreground/background clouds via nearest-neighbor box association and rotation-only background warping, then encodes them as additional conditions. The NM module adds Gaussian noise with random strength to all previous-frame latent features, making the model robust to its own errors during long-horizon inference.

What would settle it

Measure LaGen's long-horizon Chamfer distance when the conditioning boxes are provided by an off-the-shelf detector/tracker instead of ground-truth box labels; if at 5.5s–9.5s the errors are no longer clearly lower than the 6-frame-input forecasting baselines, the claim of single-frame-only generation is unsupported.

Watch

Extended reading notes

Core claim

The paper introduces LaGen, a latent-diffusion generator operating on range images that recursively produces the next LiDAR frame conditioned on the previous frame, projected 3D bounding boxes, and ego-vehicle states. Two novel modules—a Scene Decoupling Estimation (SDE) module that estimates current-frame foreground and background from the previous frame and box motion, and a Noise Modulation (NM) module that injects noise into previous-frame features to reduce training-inference mismatch—together enable long-horizon generation. Experiments on nuScenes report state-of-the-art generation (MMD 0.35e-4, JSD 3.97e-2) and strong autoregressive results, with Chamfer distance at 9.5s of 2.66 m² ve

Load-bearing premise

Per-frame 3D bounding boxes and ego-vehicle states must be supplied as conditions at every generation step; if those conditions have to be predicted rather than taken from the scene, LaGen's stated advantage over forecasting baselines likely disappears.

Editorial extensions

If this is right

  • Autonomous-driving simulators can generate realistic LiDAR sequences interactively from a single initial frame, enabling closed-loop policy evaluation.
  • Editing a bounding box at any intermediate frame changes the generated scene accordingly, including occlusions, allowing scenario-level intervention and counterfactual testing.
  • The framework generates longer horizons with relatively stable error growth, unlike forecasting models whose errors explode after a few seconds.
  • The same conditional latent-diffusion design can be adapted for data augmentation, expanding sparse real-world LiDAR collections with controlled variations.
  • Frame-by-frame generation naturally integrates per-step planner or policy decisions, a capability lacking in multi-frame-fixed prediction.

Reading between the lines

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

  • The paper's 'single-frame input' claim should be read as 'single LiDAR frame plus per-frame ground-truth bounding boxes and ego states'; a practical deployment would need these conditions predicted, a step the paper does not address.
  • A natural testable extension is to chain a 3D detector/forecaster ahead of LaGen and measure the resulting generation quality—if it degrades sharply, the advantage over prediction baselines may narrow.
  • The NM module's noise injection suggests a broader principle: autoregressive generative models should be trained under condition corruption that matches inference-time errors, a recipe transferable to other sensor modalities.
  • The SDE module's object-centric decomposition could be reused to control scene content at the instance level, which is a step toward controllable world models rather than passive simulators.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The paper proposes LaGen, an autoregressive LiDAR scene generation framework. Given a single initial LiDAR frame, LaGen recursively generates subsequent frames using a latent diffusion model conditioned on the previous frame, per-frame 3D bounding boxes, ego-vehicle states, and estimates from a proposed Scene Decoupling Estimation (SDE) module. A Noise Modulation (NM) module is added to mitigate error accumulation. The authors claim LaGen is the first framework capable of long-horizon, frame-by-frame, interactive LiDAR scene generation, and report state-of-the-art results on nuScenes for both single-frame generation and long-horizon prediction-style evaluation against 4D-Occ and ViDAR.

Significance. If validated, an autoregressive, interactive LiDAR world model would be valuable for closed-loop simulation, data augmentation, and safety evaluation. The paper introduces a plausible architecture (SDE, NM, multi-condition latent diffusion) and demonstrates object-level editing, which is a useful capability. However, the current evaluation does not establish the headline claims. The negative MMD in Table 2 is formally impossible, and the long-horizon comparison against prediction baselines is confounded by the use of ground-truth future bounding boxes and ego states as per-step conditioning. The contribution is best evaluated as a conditional LiDAR scene generator rather than a single-frame-input predictor.

major comments (5)
  1. [Table 2] Table 2 reports MMD = -0.35 (in 10^-4 units) for LaGen. Maximum Mean Discrepancy is a non-negative metric (squared RKHS distance); a negative value is impossible. This indicates an error in metric computation, sign, or reporting. The claim that LaGen outperforms LiDARGen, LiDM, and RangeLDM on MMD is unsupported until the corrected value is provided and the computation is described.
  2. [Sec. 3.3.4 and Tables 1/3] The long-horizon evaluation against 4D-Occ and ViDAR is confounded. As listed in Sec. 3.3.4, at every generation step LaGen receives B^s (3D bounding boxes and semantic labels of objects in the current frame) and E^s (current ego states). In the experiments, these are ground-truth future annotations. Prediction baselines receive only historical LiDAR, not future object layouts. Thus the large improvements in Table 1 (e.g., Chamfer 0.50 vs 1.00 at 0.5s) and Table 3 (L1 0.13 vs 1.23 at 0.5s) may largely reflect oracle information rather than superior generative or predictive ability. The comparison is only meaningful if baselines receive the same conditioning, or if B^s and E^s are first predicted from the single-frame input and the prediction error is propagated.
  3. [Abstract and Fig. 1] The claim that LaGen works "based solely on single-frame input" is contradicted by the method description in Sec. 3.2 and Sec. 3.3.4, where per-frame 3D bounding boxes and ego states are required conditioning inputs at each autoregressive step. In the nuScenes experiments, these are ground-truth annotations for the future time steps. The paper should clearly distinguish between (a) unconditional single-frame-to-sequence prediction and (b) interactive generation conditioned on per-frame tracked scene layout. As written, the abstract and Figure 1 overstate the input requirement and the comparison to prediction baselines.
  4. [Sec. 4.3.1] The evaluation protocol is underspecified. The paper states that 10-second segments are kept only if all frames belong to the same scene, yielding 176 valid scenes, but does not report the total number of segments before filtering, the selection criteria (e.g., scenes with no ego stops), or whether baselines are retrained on this subset. The 'rolling inference' extension of 4D-Occ is described in one sentence; this extension may significantly affect the baseline's error accumulation and should be detailed and released with code.
  5. [Sec. 3.3.2 and Sec. 3.3.4] It is unclear whether the SDE module uses ground-truth bounding boxes during inference. Eq. (9) uses C_s^{ij}, the centers of current-frame bounding boxes, which in the experiments are ground-truth current-frame boxes. Sec. 3.3.4 says SDE is computed from generated data, but B^s is still listed as an input. If ground-truth boxes are used at inference, the SDE is an oracle component for object positions; if they are not, the equations and input list must be revised. This distinction is central to assessing the interactive-generation claim in Sec. 4.4, where editing is performed by changing ground-truth boxes.
minor comments (6)
  1. [Sec. 1 and Sec. 4.1] Typos: 'Dateset' should be 'Dataset'; 'predicte' should be 'predict'.
  2. [Table 2 header] 'Y ears' should be 'Years'.
  3. [Sec. 3.1 and Tables 1/3] Units are inconsistent: Table 1 lists Chamfer Distance in m^2 while Table 3 lists L1 error in m. Clarify whether the same metric is used, and define the ray-depth metrics precisely (which rays, how depth is computed, how point sparsity is handled).
  4. [Sec. 3.3.3] The noise modulation hyperparameter N is not specified. Since the NM module is a key contribution, report the value or range of N used in training and inference.
  5. [Fig. 3 and Sec. 3.3.1] In the cross-attention formulation of Eq. (6), q=H_B, k=H_B^prev, v=H_B^cur is asymmetric: the query comes from current-box features while keys come from previous features. The purpose of this mixing should be explained in the text.
  6. [Sec. 4.2] The generator evaluation in Table 2 uses MMD and JSD following LiDARGen, but the paper does not state how many samples are generated, whether the same seeds are used, or whether the MMD is computed in point-cloud space or range-image space. This is needed for reproducibility.

Circularity Check

2 steps flagged · score 6.0 of 10

Oracle conditioning of current-frame GT boxes and ego states makes the long-horizon 'prediction' comparison partly circular.

  1. self definitional [Sec. 3.3.2, Eq. (9); Sec. 3.3.4; Tables 1/3]
    "To estimate the position of \hat p_{ij} in the current-frame scene, we further design a nearest neighbor search algorithm targeting the centers of bounding boxes of the same category. ... we obtain an estimated foreground object point cloud p^s_{ij} for the current frame by: e p^s_{ij} = \hat p^{s-1}_{ij} + (C^s_{ij} - C^{s-1}_{ij'})."

    Eq. (9) constructs the estimated current-frame foreground by translating the previous frame's object points with the displacement between ground-truth current and previous bounding-box centers. Since B^s (current-frame GT boxes) is listed as a per-step input in Sec. 3.3.4, the object-location part of the generated frame is provided by the target frame itself. The Chamfer/L1 errors reported against future GT frames in Tables 1/3 therefore measure rendering of known layouts, not prediction of them; the large gap over 4D-Occ/ViDAR, which receive no future boxes, is partly forced by construction.

  2. other [Sec. 3.2 vs Abstract/Fig. 1; Sec. 3.3.4]
    "Accordingly, at each step of the generation process, we aim to generate the current LiDAR frame based solely on the previous frame, while incorporating 3D bounding box information and ego-vehicle states from the previous and current frames."

    The paper's central headline ('based solely on single-frame input') is contradicted by its own generation equation: every step takes the current frame's boxes B^s and ego states E^s as conditions. Those are future information for that step, so the 'long-horizon scene generation' result is derived from the per-frame target layout, not from a single starting frame. This makes the single-frame-input claim and the resulting SOTA comparison circular rather than first-principles.

full rationale

The autoregressive architecture itself has genuine content: it synthesizes point-level geometry with a latent diffusion U-Net, and the NM/SDE modules are not vacuous. The output is not literally the input boxes. However, the benchmark claim that LaGen 'predicts' long-horizon scenes from one frame is not self-contained. Every autoregressive step consumes the current frame's ground-truth 3D boxes and ego states (Sec. 3.3.4), and Eq. (9) directly places the foreground using current GT box centers, so the object-layout part of the generated frame is oracle-supplied rather than predicted. The comparisons to 4D-Occ and ViDAR, which receive only historical point clouds, therefore do not establish the claimed single-frame-input advantage. This is partial circularity, not full: the model still must generate realistic point distributions and background geometry, and the claimed architectural novelty is not disproved. I also flag a separate numerical red flag: Table 2 reports MMD = -0.35 for LaGen, an impossible value for a valid MMD, so that table cannot independently support the generation-quality claims. No load-bearing self-citation chain or imported uniqueness theorem was found.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper does not postulate new physical entities. The two new modules (SDE and NM) are network components, not entities. The main dependencies are the conditional inputs (bounding boxes, ego states) and the range-image inversion, which are domain-specific assumptions. One free hyperparameter (N) is not even reported, and the benchmark filter is a data-selection parameter.

free parameters (3)
  • Noise modulation noise level N
    Equation (11) samples n uniformly from [0, N], but N is never specified in the paper. This is a hyperparameter controlling how much noise is injected into the previous frame's latent features; the choice plausibly trades-off short-horizon accuracy against long-horizon stability.
  • 10-second scene segment filter (176 valid scenes) = 176
    The evaluation protocol removes any 10-second segment that 'contains frames belonging to different scenes', leaving only 176 valid scenes from the nuScenes dataset. The filter is a data-processing choice that may bias the benchmark and is not justified with statistics.
  • Diffusion and UNet hyperparameters
    The number of diffusion timesteps, latent dimensionality, UNet base channels, and training schedule are not listed in the paper. These are standard but non-trivial choices that are needed to reproduce the results; they are only implied by the reference to RangeLDM.
assumptions (4)
  • domain assumption Spherical range-image representation is invertible and preserves scene geometry (Eq. 1-2).
    The method converts 3D point clouds to 2D range images, assumes this projection loses no important information, and later reverse-projects to 3D. The validity depends on the sensor model and the Hough-voting estimates of h_j and phi_j, inherited from RangeLDM.
  • domain assumption Points inside the 3D bounding boxes are 'foreground' and all other points are 'background' (Eq. 7).
    The SDE module decouples the scene based on bounding box membership. This is a simplifying approximation: a point that is inside a bounding box belongs to the object, and everything else is static background, which may ignore occlusions, sensor noise, and unlabeled objects.
  • ad hoc to paper The current frame's background can be estimated from the previous frame by applying only the rotation component of the relative pose (Eq. 10).
    Equation (10) transforms the previous background using only E^{s-1}_rotate, ignoring translation. For a moving ego-vehicle, the background changes with translation as well, so this is a crude approximation that the model is expected to correct.
  • domain assumption 3D bounding boxes for the current frame are available at inference time.
    Section 3.3.4 lists B^s as a required input. This makes the generator conditional on the future object layout, which is only available in a simulator or with an upstream perception/prediction module, and is not available from a single input frame alone.

how reviews work

0 comments
Cite this review

Pith. "Pith review of LaGen: Towards Autoregressive LiDAR Scene Generation." pith.science (2026). https://pith.science/paper/23G7G2E3

@misc{pith2026251121256,
  author       = {Pith},
  title        = {Pith review of: LaGen: Towards Autoregressive LiDAR Scene Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/23G7G2E3}},
  note         = {Machine review of arXiv:2511.21256}
}
read the original abstract

Generative world models for autonomous driving (AD) are of great value in applications such as data augmentation, closed-loop simulation, and safety-critical scenario evaluation. Unlike the widely studied image modality, in this work we explore generative world models for LiDAR data. Existing generation methods for LiDAR predominantly focus on single frame generation or lack the capacity for interactive simulation, while existing prediction approaches require multiple frames of historical input and can only deterministically predict multiple frames at once. Both paradigms fail to support long-horizon interactive generation. To this end, we introduce \textbf{LaGen}, which, to the best of our knowledge is the first autoregressive framework capable of generating long-horizon LiDAR scenes in a frame-by-frame, interactive manner. LaGen is able to take a single-frame input as a starting point and effectively utilize bounding box information as conditions to generate high-fidelity 4D scene. In addition, we introduce a scene decoupling estimation module to enhance the model's interactive generation capability for object-level content, as well as a noise modulation module to mitigate error accumulation during long-horizon generation. We extensively evaluate LaGen's performance in controlled data generation and long-horizon scene generation on the nuScenes dataset. The experimental results demonstrate that LaGen achieves state-of-the-art performance, especially on later frames. The code is publicly available at: https://github.com/szzhou88/LaGen.

Figures

Figures reproduced from arXiv: 2511.21256 by the authors.

Figure 1
Figure 1. This work introduces LaGen, a novel 4D LiDAR scene generation framework that can autoregressively generate [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of the LaGen framework. It first obtains estimates of the current frame’s foreground and background point clouds via the Scene Decoupling Estimation (SDE) module. Subsequently, all three-dimensional information is projected to the two-dimensional range image space via spherical mapping. These range-view representations are then processed by a pretrained VAE and a noise modulation (NM) module, before being p… view at source ↗
Figure 3
Figure 3. The schematic diagram of the LiDAR generator archi [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (3 more)
Figure 5
Figure 5. Figure 5: The schematic diagram of noise modulation module. [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Qualitative results. LaGen, based on single-frame input, clearly outperforms other baseline methods in prediction accuracy. the rotation matrix: Pes bg = trans(P s−1 bg , Es−1 rotate). (10) Finally, we project both Pes obj and Pes bg into the range-view space and encod…
Figure 7
Figure 7. Figure 7: Object-level editing and interaction. By moving or [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

70 extracted references · 10 linked inside Pith

  1. [1]

    Deep generative modeling of lidar data

    Lucas Caccia, Herke Van Hoof, Aaron Courville, and Joelle Pineau. Deep generative modeling of lidar data. In2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 5034–5040. IEEE, 2019. 2

    2019

  2. [2]

    nuscenes: A multi- modal dataset for autonomous driving

    Holger Caesar, Varun Bankiti, Alex H Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Gi- ancarlo Baldan, and Oscar Beijbom. nuscenes: A multi- modal dataset for autonomous driving. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11621–11631, 2020. 2, 7

    2020

  3. [3]

    Diffusion forcing: Next-token prediction meets full-sequence diffu- sion.Advances in Neural Information Processing Systems, 37:24081–24125, 2024

    Boyuan Chen, Diego Mart ´ı Mons ´o, Yilun Du, Max Sim- chowitz, Russ Tedrake, and Vincent Sitzmann. Diffusion forcing: Next-token prediction meets full-sequence diffu- sion.Advances in Neural Information Processing Systems, 37:24081–24125, 2024. 6

    2024

  4. [4]

    Language-guided 3d object detection in point cloud for autonomous driving.arXiv preprint arXiv:2305.15765, 2023

    Wenhao Cheng, Junbo Yin, Wei Li, Ruigang Yang, and Jianbing Shen. Language-guided 3d object detection in point cloud for autonomous driving.arXiv preprint arXiv:2305.15765, 2023. 2

    arXiv 2023

  5. [5]

    Transfuser: Imitation with transformer-based sensor fusion for autonomous driv- ing.IEEE transactions on pattern analysis and machine in- telligence, 45:12878–12895, 2022

    Kashyap Chitta, Aditya Prakash, Bernhard Jaeger, Zehao Yu, Katrin Renz, and Andreas Geiger. Transfuser: Imitation with transformer-based sensor fusion for autonomous driv- ing.IEEE transactions on pattern analysis and machine in- telligence, 45:12878–12895, 2022. 2

    2022

  6. [6]

    Openscene: The largest up-to-date 3d occupancy prediction benchmark in autonomous driv- ing

    OpenScene Contributors. Openscene: The largest up-to-date 3d occupancy prediction benchmark in autonomous driv- ing. InProceedings of the Conference on Computer Vision and Pattern Recognition, Vancouver, Canada, pages 18–22,

  7. [7]

    Streetscapes: Large-scale consistent street view generation using autore- gressive video diffusion

    Boyang Deng, Richard Tucker, Zhengqi Li, Leonidas Guibas, Noah Snavely, and Gordon Wetzstein. Streetscapes: Large-scale consistent street view generation using autore- gressive video diffusion. InACM SIGGRAPH 2024 Confer- ence Papers, pages 1–11, 2024. 6

    2024

  8. [8]

    Carla: An open urban driv- ing simulator

    Alexey Dosovitskiy, German Ros, Felipe Codevilla, Anto- nio Lopez, and Vladlen Koltun. Carla: An open urban driv- ing simulator. InConference on robot learning, pages 1–16. PMLR, 2017. 2

    2017

Show all 70 references
  1. [9]

    A point set generation network for 3d object reconstruction from a single image

    Haoqiang Fan, Hao Su, and Leonidas J Guibas. A point set generation network for 3d object reconstruction from a single image. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 605–613, 2017. 7

    2017

  2. [10]

    Vision meets robotics: The KITTI dataset.Inter- national Journal of Robotics Research, 32(11):1231 – 1237,

    Andreas Geiger, Philip Lenz, Christoph Stiller, and Raquel Urtasun. Vision meets robotics: The KITTI dataset.Inter- national Journal of Robotics Research, 32(11):1231 – 1237,

  3. [11]

    Generative adversarial nets.Advances in neural information processing systems, 27, 2014

    Ian J Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets.Advances in neural information processing systems, 27, 2014. 2, 3

    2014

  4. [12]

    Vip3d: End-to-end visual trajectory prediction via 3d agent queries

    Junru Gu, Chenxu Hu, Tianyuan Zhang, Xuanyao Chen, Yilun Wang, Yue Wang, and Hang Zhao. Vip3d: End-to-end visual trajectory prediction via 3d agent queries. InProceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 5496–5506, 2023. 7

    2023

  5. [13]

    Denoising dif- fusion probabilistic models.Advances in neural information processing systems, 33:6840–6851, 2020

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising dif- fusion probabilistic models.Advances in neural information processing systems, 33:6840–6851, 2020. 2

    2020

  6. [14]

    Monocular quasi-dense 3d object tracking.IEEE Transactions on Pattern Analysis and Machine Intelligence, 45:1992–2008, 2022

    Hou-Ning Hu, Yung-Hsu Yang, Tobias Fischer, Trevor Dar- rell, Fisher Yu, and Min Sun. Monocular quasi-dense 3d object tracking.IEEE Transactions on Pattern Analysis and Machine Intelligence, 45:1992–2008, 2022. 7

    1992

  7. [15]

    Rangeldm: Fast realistic lidar point cloud generation

    Qianjiang Hu, Zhimin Zhang, and Wei Hu. Rangeldm: Fast realistic lidar point cloud generation. InEuropean Confer- ence on Computer Vision, pages 115–135. Springer, 2024. 3, 7

    2024

  8. [16]

    St-p3: End-to-end vision-based au- tonomous driving via spatial-temporal feature learning

    Shengchao Hu, Li Chen, Penghao Wu, Hongyang Li, Junchi Yan, and Dacheng Tao. St-p3: End-to-end vision-based au- tonomous driving via spatial-temporal feature learning. In European Conference on Computer Vision, pages 533–549. Springer, 2022. 7

    2022

  9. [17]

    Planning-oriented autonomous driving

    Yihan Hu, Jiazhi Yang, Li Chen, Keyu Li, Chonghao Sima, Xizhou Zhu, Siqi Chai, Senyao Du, Tianwei Lin, Wenhai Wang, et al. Planning-oriented autonomous driving. InPro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 17853–17862, 2023. 2, 7

    2023

  10. [18]

    Bevdet: High-performance multi-camera 3d object de- tection in bird-eye-view.arXiv preprint arXiv:2112.11790,

    Junjie Huang, Guan Huang, Zheng Zhu, Yun Ye, and Dalong Du. Bevdet: High-performance multi-camera 3d object de- tection in bird-eye-view.arXiv preprint arXiv:2112.11790,

    arXiv

  11. [19]

    Bench2drive: Towards multi-ability bench- marking of closed-loop end-to-end autonomous driving.Ad- vances in Neural Information Processing Systems, 37:819– 844, 2024

    Xiaosong Jia, Zhenjie Yang, Qifeng Li, Zhiyuan Zhang, and Junchi Yan. Bench2drive: Towards multi-ability bench- marking of closed-loop end-to-end autonomous driving.Ad- vances in Neural Information Processing Systems, 37:819– 844, 2024. 2

    2024

  12. [20]

    Vad: Vectorized scene representa- tion for efficient autonomous driving

    Bo Jiang, Shaoyu Chen, Qing Xu, Bencheng Liao, Jiajie Chen, Helong Zhou, Qian Zhang, Wenyu Liu, Chang Huang, and Xinggang Wang. Vad: Vectorized scene representa- tion for efficient autonomous driving. InProceedings of the IEEE/CVF International Conference on Computer Vision, p...

    2023

  13. [21]

    Towards learning-based planning: 9 The nuplan benchmark for real-world autonomous driving

    Napat Karnchanachari, Dimitris Geromichalos, Kok Seang Tan, Nanxiang Li, Christopher Eriksen, Shakiba Yaghoubi, Noushin Mehdipour, Gianmarco Bernasconi, Whye Kit Fong, Yiluan Guo, et al. Towards learning-based planning: 9 The nuplan benchmark for real-world autonomous driving....

    2024

  14. [22]

    Point cloud forecasting as a proxy for 4d occupancy forecasting

    Tarasha Khurana, Peiyun Hu, David Held, and Deva Ra- manan. Point cloud forecasting as a proxy for 4d occupancy forecasting. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 1116– 1124, 2023. 2, 3, 7, 8

    2023

  15. [23]

    Variational diffusion models.Advances in neural infor- mation processing systems, 34:21696–21707, 2021

    Diederik Kingma, Tim Salimans, Ben Poole, and Jonathan Ho. Variational diffusion models.Advances in neural infor- mation processing systems, 34:21696–21707, 2021. 2

    2021

  16. [24]

    Auto-encoding varia- tional bayes.arXiv preprint arXiv:1312.6114, 2013

    Diederik P Kingma and Max Welling. Auto-encoding varia- tional bayes.arXiv preprint arXiv:1312.6114, 2013. 2

    2013 arXiv

  17. [25]

    Lwsis: Lidar-guided weakly supervised instance segmentation for autonomous driving

    Xiang Li, Junbo Yin, Botian Shi, Yikang Li, Ruigang Yang, and Jianbing Shen. Lwsis: Lidar-guided weakly supervised instance segmentation for autonomous driving. InProceed- ings of the AAAI conference on artificial intelligence, pages 1433–1441, 2023. 2

    2023

  18. [26]

    Di-v2x: Learning domain- invariant representation for vehicle-infrastructure collabora- tive 3d object detection

    Xiang Li, Junbo Yin, Wei Li, Chengzhong Xu, Ruigang Yang, and Jianbing Shen. Di-v2x: Learning domain- invariant representation for vehicle-infrastructure collabora- tive 3d object detection. InProceedings of the AAAI Confer- ence on Artificial Intelligence, pages 3208–3215, 2024. 2

    2024

  19. [27]

    Bevdepth: Acquisition of reliable depth for multi-view 3d object detec- tion

    Yinhao Li, Zheng Ge, Guanyi Yu, Jinrong Yang, Zengran Wang, Yukang Shi, Jianjian Sun, and Zeming Li. Bevdepth: Acquisition of reliable depth for multi-view 3d object detec- tion. InProceedings of the AAAI conference on artificial intelligence, pages 1477–1485, 2023. 7

    2023

  20. [28]

    End-to-end 3d tracking with decoupled queries

    Yanwei Li, Zhiding Yu, Jonah Philion, Anima Anandku- mar, Sanja Fidler, Jiaya Jia, and Jose Alvarez. End-to-end 3d tracking with decoupled queries. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 18302–18311, 2023

    2023

  21. [29]

    Bevformer: learning bird’s-eye-view representation from lidar-camera via spatiotemporal transformers.IEEE Transactions on Pat- tern Analysis and Machine Intelligence, 2024

    Zhiqi Li, Wenhai Wang, Hongyang Li, Enze Xie, Chong- hao Sima, Tong Lu, Qiao Yu, and Jifeng Dai. Bevformer: learning bird’s-eye-view representation from lidar-camera via spatiotemporal transformers.IEEE Transactions on Pat- tern Analysis and Machine Intelligence, 2024. 7

    2024

  22. [30]

    Pnpnet: End-to-end per- ception and prediction with tracking in the loop

    Ming Liang, Bin Yang, Wenyuan Zeng, Yun Chen, Rui Hu, Sergio Casas, and Raquel Urtasun. Pnpnet: End-to-end per- ception and prediction with tracking in the loop. InProceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 11553–11562, 2020. 7

    2020

  23. [31]

    Flownet3d: Learning scene flow in 3d point clouds

    Xingyu Liu, Charles R Qi, and Leonidas J Guibas. Flownet3d: Learning scene flow in 3d point clouds. InPro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 529–537, 2019. 3

    2019

  24. [32]

    Pcpnet: An efficient and semantic-enhanced transformer net- work for point cloud prediction.IEEE Robotics and Automa- tion Letters, 8(7):4267–4274, 2023

    Zhen Luo, Junyi Ma, Zijie Zhou, and Guangming Xiong. Pcpnet: An efficient and semantic-enhanced transformer net- work for point cloud prediction.IEEE Robotics and Automa- tion Letters, 8(7):4267–4274, 2023. 3

    2023

  25. [33]

    Lidar- only based navigation algorithm for an autonomous agricul- tural robot.Computers and electronics in agriculture, 154: 71–79, 2018

    Flavio BP Malavazi, Remy Guyonneau, Jean-Baptiste Fasquel, Sebastien Lagrange, and Franck Mercier. Lidar- only based navigation algorithm for an autonomous agricul- tural robot.Computers and electronics in agriculture, 154: 71–79, 2018. 2

    2018

  26. [34]

    Weakly supervised 3d object detection from lidar point cloud

    Qinghao Meng, Wenguan Wang, Tianfei Zhou, Jianbing Shen, Luc Van Gool, and Dengxin Dai. Weakly supervised 3d object detection from lidar point cloud. InEuropean Con- ference on computer vision, pages 515–531. Springer, 2020. 2

    2020

  27. [35]

    Self-supervised point cloud prediction using 3d spatio-temporal convolutional networks

    Benedikt Mersch, Xieyuanli Chen, Jens Behley, and Cyrill Stachniss. Self-supervised point cloud prediction using 3d spatio-temporal convolutional networks. InConference on Robot Learning, pages 1444–1454. PMLR, 2022. 3

    2022

  28. [36]

    Lidar data synthe- sis with denoising diffusion probabilistic models

    Kazuto Nakashima and Ryo Kurazume. Lidar data synthe- sis with denoising diffusion probabilistic models. In2024 IEEE International Conference on Robotics and Automation (ICRA), pages 14724–14731. IEEE, 2024. 2, 3

    2024

  29. [37]

    Glide: Towards photorealistic image generation and editing with text-guided diffusion models.arXiv preprint arXiv:2112.10741, 2021

    Alex Nichol, Prafulla Dhariwal, Aditya Ramesh, Pranav Shyam, Pamela Mishkin, Bob McGrew, Ilya Sutskever, and Mark Chen. Glide: Towards photorealistic image generation and editing with text-guided diffusion models.arXiv preprint arXiv:2112.10741, 2021. 2

    2021 arXiv

  30. [38]

    Improved denoising diffusion probabilistic models

    Alexander Quinn Nichol and Prafulla Dhariwal. Improved denoising diffusion probabilistic models. InInternational conference on machine learning, pages 8162–8171. PMLR,

  31. [39]

    Atppnet: Attention based temporal point cloud prediction network

    Kaustab Pal, Aditya Sharma, Avinash Sharma, and K Mad- hava Krishna. Atppnet: Attention based temporal point cloud prediction network. In2024 IEEE International Conference on Robotics and Automation (ICRA), pages 11140–11146. IEEE, 2024. 3

    2024

  32. [40]

    Simpletrack: Understanding and rethinking 3d multi-object tracking

    Ziqi Pang, Zhichao Li, and Naiyan Wang. Simpletrack: Understanding and rethinking 3d multi-object tracking. In European Conference on Computer Vision, pages 680–696. Springer, 2022. 7

    2022

  33. [41]

    Multi- modal fusion transformer for end-to-end autonomous driv- ing

    Aditya Prakash, Kashyap Chitta, and Andreas Geiger. Multi- modal fusion transformer for end-to-end autonomous driv- ing. InProceedings of the IEEE/CVF conference on com- puter vision and pattern recognition, pages 7077–7087,

  34. [42]

    Hierarchical text-conditional image gener- ation with clip latents.arXiv preprint arXiv:2204.06125, 1 (2):3, 2022

    Aditya Ramesh, Prafulla Dhariwal, Alex Nichol, Casey Chu, and Mark Chen. Hierarchical text-conditional image gener- ation with clip latents.arXiv preprint arXiv:2204.06125, 1 (2):3, 2022. 2

    2022 arXiv

  35. [43]

    Towards realistic scene generation with lidar diffusion models

    Haoxi Ran, Vitor Guizilini, and Yue Wang. Towards realistic scene generation with lidar diffusion models. InProceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 14738–14748, 2024. 3, 7

    2024

  36. [44]

    Drone laser scanning for modeling riverscape topography and vegetation: Comparison with traditional aerial lidar

    Jonathan P Resop, Laura Lehmann, and W Cully Hession. Drone laser scanning for modeling riverscape topography and vegetation: Comparison with traditional aerial lidar. Drones, 3(2):35, 2019. 2

    2019

  37. [45]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj ¨orn Ommer. High-resolution image synthesis with latent diffusion models. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684–10695, 2022. 2, 3

    2022

  38. [46]

    Photorealistic text-to-image diffusion models with deep language understanding.Advances in neural information processing systems, 35:36479–36494, 2022

    Chitwan Saharia, William Chan, Saurabh Saxena, Lala Li, Jay Whang, Emily L Denton, Kamyar Ghasemipour, Raphael Gontijo Lopes, Burcu Karagol Ayan, Tim Salimans, et al. Photorealistic text-to-image diffusion models with deep language understanding.Advances in neural information ...

    2022

  39. [47]

    Pointr- cnn: 3d object proposal generation and detection from point cloud

    Shaoshuai Shi, Xiaogang Wang, and Hongsheng Li. Pointr- cnn: 3d object proposal generation and detection from point cloud. InProceedings of the IEEE/CVF conference on com- puter vision and pattern recognition, pages 770–779, 2019. 2

    2019

  40. [48]

    Pv-rcnn: Point- voxel feature set abstraction for 3d object detection

    Shaoshuai Shi, Chaoxu Guo, Li Jiang, Zhe Wang, Jianping Shi, Xiaogang Wang, and Hongsheng Li. Pv-rcnn: Point- voxel feature set abstraction for 3d object detection. InPro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10529–10538, 2020. 2

    2020

  41. [49]

    Deep unsupervised learning using nonequilibrium thermodynamics

    Jascha Sohl-Dickstein, Eric Weiss, Niru Maheswaranathan, and Surya Ganguli. Deep unsupervised learning using nonequilibrium thermodynamics. InInternational confer- ence on machine learning, pages 2256–2265. pmlr, 2015. 2

    2015

  42. [50]

    Denoising diffusion implicit models.arXiv preprint arXiv:2010.02502, 2020

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denoising diffusion implicit models.arXiv preprint arXiv:2010.02502, 2020. 3

    2010 arXiv

  43. [51]

    Generative modeling by esti- mating gradients of the data distribution.Advances in neural information processing systems, 32, 2019

    Yang Song and Stefano Ermon. Generative modeling by esti- mating gradients of the data distribution.Advances in neural information processing systems, 32, 2019. 2

    2019

  44. [52]

    Improved techniques for training score-based generative models.Advances in neural information processing systems, 33:12438–12448, 2020

    Yang Song and Stefano Ermon. Improved techniques for training score-based generative models.Advances in neural information processing systems, 33:12438–12448, 2020

    2020

  45. [53]

    Score-based generative modeling through stochastic differential equa- tions.arXiv preprint arXiv:2011.13456, 2020

    Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Ab- hishek Kumar, Stefano Ermon, and Ben Poole. Score-based generative modeling through stochastic differential equa- tions.arXiv preprint arXiv:2011.13456, 2020. 2

    2011 arXiv

  46. [54]

    Scene as occupancy

    Wenwen Tong, Chonghao Sima, Tai Wang, Li Chen, Silei Wu, Hanming Deng, Yi Gu, Lewei Lu, Ping Luo, Dahua Lin, et al. Scene as occupancy. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 8406– 8415, 2023. 7

    2023

  47. [55]

    Plant detection and mapping for agricultural robots using a 3d lidar sensor.Robotics and autonomous systems, 59(5):265–273, 2011

    Ulrich Weiss and Peter Biber. Plant detection and mapping for agricultural robots using a 3d lidar sensor.Robotics and autonomous systems, 59(5):265–273, 2011. 2

    2011

  48. [56]

    Inverting the pose forecasting pipeline with spf2: Sequential pointcloud forecasting for se- quential pose forecasting

    Xinshuo Weng, Jianren Wang, Sergey Levine, Kris Kitani, and Nicholas Rhinehart. Inverting the pose forecasting pipeline with spf2: Sequential pointcloud forecasting for se- quential pose forecasting. InConference on robot learning, pages 11–20. PMLR, 2021. 2, 3

    2021

  49. [57]

    S2net: Stochastic sequential pointcloud forecasting

    Xinshuo Weng, Junyu Nan, Kuan-Hui Lee, Rowan McAl- lister, Adrien Gaidon, Nicholas Rhinehart, and Kris M Ki- tani. S2net: Stochastic sequential pointcloud forecasting. In European Conference on Computer Vision, pages 549–564. Springer, 2022. 2, 3

    2022

  50. [58]

    Para-drive: Parallelized architecture for real- time autonomous driving

    Xinshuo Weng, Boris Ivanovic, Yan Wang, Yue Wang, and Marco Pavone. Para-drive: Parallelized architecture for real- time autonomous driving. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 15449–15458, 2024. 2

    2024

  51. [59]

    Argoverse 2: Next generation datasets for self-driving perception and forecasting.arXiv preprint arXiv:2301.00493, 2023

    Benjamin Wilson, William Qi, Tanmay Agarwal, John Lambert, Jagjeet Singh, Siddhesh Khandelwal, Bowen Pan, Ratnesh Kumar, Andrew Hartnett, Jhony Kaesemodel Pontes, et al. Argoverse 2: Next generation datasets for self-driving perception and forecasting.arXiv preprint arXiv:2301...

    2023 arXiv

  52. [60]

    Deep 3d object detection networks using lidar data: A review.IEEE Sensors Journal, 21(2):1152–1171, 2020

    Yutian Wu, Yueyu Wang, Shuwei Zhang, and Harutoshi Ogai. Deep 3d object detection networks using lidar data: A review.IEEE Sensors Journal, 21(2):1152–1171, 2020. 2

    2020

  53. [61]

    Ultralidar: Learning compact representations for lidar completion and generation.arXiv preprint arXiv:2311.01448, 2023

    Yuwen Xiong, Wei-Chiu Ma, Jingkang Wang, and Raquel Urtasun. Ultralidar: Learning compact representations for lidar completion and generation.arXiv preprint arXiv:2311.01448, 2023. 2, 3

    2023 arXiv

  54. [62]

    Second: Sparsely embed- ded convolutional detection.Sensors, 18(10), 2018

    Yan Yan, Yuxing Mao, and Bo Li. Second: Sparsely embed- ded convolutional detection.Sensors, 18(10), 2018. 2

    2018

  55. [63]

    Visual point cloud forecasting enables scalable autonomous driving

    Zetong Yang, Li Chen, Yanan Sun, and Hongyang Li. Visual point cloud forecasting enables scalable autonomous driving. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 14673–14684, 2024. 3, 7, 8

    2024

  56. [64]

    Is-fusion: Instance-scene collaborative fusion for multimodal 3d ob- ject detection

    Junbo Yin, Jianbing Shen, Runnan Chen, Wei Li, Ruigang Yang, Pascal Frossard, and Wenguan Wang. Is-fusion: Instance-scene collaborative fusion for multimodal 3d ob- ject detection. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14905...

    2024

  57. [65]

    Vector-quantized image modeling with improved vqgan.arXiv preprint arXiv:2110.04627, 2021

    Jiahui Yu, Xin Li, Jing Yu Koh, Han Zhang, Ruoming Pang, James Qin, Alexander Ku, Yuanzhong Xu, Jason Baldridge, and Yonghui Wu. Vector-quantized image modeling with improved vqgan.arXiv preprint arXiv:2110.04627, 2021. 3

    2021 arXiv

  58. [66]

    Optical flow and scene flow estimation: A survey

    Mingliang Zhai, Xuezhi Xiang, Ning Lv, and Xiangdong Kong. Optical flow and scene flow estimation: A survey. Pattern Recognition, 114:107861, 2021. 3

    2021

  59. [67]

    Loam: Lidar odometry and mapping in real-time

    Ji Zhang, Sanjiv Singh, et al. Loam: Lidar odometry and mapping in real-time. InRobotics: Science and systems, pages 1–9. Berkeley, CA, 2014. 2

    2014

  60. [68]

    Adding conditional control to text-to-image diffusion models

    Lvmin Zhang, Anyi Rao, and Maneesh Agrawala. Adding conditional control to text-to-image diffusion models. In Proceedings of the IEEE/CVF international conference on computer vision, pages 3836–3847, 2023. 2

    2023

  61. [69]

    Mutr3d: A multi-camera tracking frame- work via 3d-to-2d queries

    Tianyuan Zhang, Xuanyao Chen, Yue Wang, Yilun Wang, and Hang Zhao. Mutr3d: A multi-camera tracking frame- work via 3d-to-2d queries. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4537–4546, 2022. 7

    2022

  62. [70]

    Learning to generate realistic lidar point clouds

    Vlas Zyrianov, Xiyue Zhu, and Shenlong Wang. Learning to generate realistic lidar point clouds. InEuropean Conference on Computer Vision, pages 17–35. Springer, 2022. 2, 3, 7 11

    2022

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.