Pith. sign in

REVIEW 4 major objections 6 minor 43 references

SliceSemOcc: Vertical Slice Based Multimodal 3D Semantic Occupancy Representation

T0 review · 4 major / 6 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read SliceSemOcc claims that splitting voxel features into global and local vertical slices, fused by cross attention with per-height channel weights, improves 3D semantic occupancy mIoU by 14.2% on SurroundOcc and 13.9% on OpenOccupancy, with t

desk verdict Solid height-aware attention idea, but the reported gains over M-CONet are inflated by an uncontrolled baseline; the real internal gain is about 1.2 mIoU. read the letter →

arxiv 2509.03999 v1 pith:OM5OAWUW submitted 2025-09-04 cs.CV

Han Huang , Han Sun , Ningzhong Liu , Huiyu Zhou , Jiaquan Shen This is my paper · ORCID
classification cs.CV
keywords 3DsemanticoccupancypredictionverticalslicingchannelattentionmultimodalfusionLiDAR-cameraautonomousdrivingnuScenesvoxelrepresentation
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 tries to establish that 3D semantic occupancy prediction improves when the network treats the height axis explicitly instead of pooling over the whole voxel volume. It proposes two mechanisms: global and local vertical slices of the voxel features, fused by bidirectional cross attention, and a channel-attention variant that gives each height layer its own channel weights. If correct, the framework offers a modular way to get materially better voxel-level semantic labels, especially for small objects such as pedestrians, bicycles, and traffic cones that occupy narrow height bands. The claim is supported by mIoU gains on two nuScenes-based benchmarks and by ablations isolating each design choice.

What carries the argument

The Vertical Slice Fusion (VSF) module and the SEAttention3D module. VSF splits the voxel volume along height into one global slice spanning [-5,3] m and six local slices concentrated where small objects sit, processes each with SEAttention3D, then fuses global and local features through bidirectional cross attention. SEAttention3D replaces SENet-style global pooling with pooling over only the X-Y plane, so it produces one channel-attention vector per height layer instead of one per volume, letting the network weight the same channel differently at different heights.

What would settle it

A decisive check: recompute the slice boundaries from the training split's height histogram only (or use uniform bins) and evaluate on a dataset with a different height profile; if the mIoU advantage over M-CONet collapses or reverses, the vertical-slice benefit is dataset-specific. The paper's own Table 4 gives the comparison point: hand-picked bins score 28.21 mIoU versus 27.67 for uniform 1-meter bins, so the open question is whether that gap survives train-only boundary selection and cross-dataset transfer.

Watch

Extended reading notes

Core claim

The paper reports that processing 3D occupancy voxel features with vertical slices rather than treating the volume uniformly improves semantic occupancy prediction. On nuScenes-SurroundOcc, the full SliceSemOcc pipeline raises mean IoU from 24.7% for the M-CONet baseline to 28.2% (a 14.2% relative gain), with the largest per-class gains on small objects: bicycle +53.1%, motorcycle +37.2%, pedestrian +30.0%, barrier +21.3%, traffic cone +20.4%. On nuScenes-OpenOccupancy mIoU rises from 20.1% to 22.9% (a 13.9% relative gain). Ablations attribute the gains to the combination of global and local vertical slices, to choosing local slice boundaries from the class height distribution rather than un

Load-bearing premise

The local slice boundaries were picked by looking at the height distribution of objects in the same dataset the method is then evaluated on, so the reported gains may partly reflect tuning to that dataset's statistics rather than a general principle.

Editorial extensions

If this is right

  • On nuScenes-SurroundOcc, SliceSemOcc raises mIoU from 24.7% (M-CONet) to 28.2%, a relative gain of 14.2%.
  • Gains are concentrated on small objects: barrier, bicycle, motorcycle, pedestrian, and traffic cone each improve by 20% to 53% relative.
  • On nuScenes-OpenOccupancy, mIoU rises from 20.1% to 22.9%, a relative gain of 13.9%.
  • Ablations show that height-distribution-based local bins beat uniform bins (28.21 vs 27.67 mIoU), and cross-attention fusion beats simple concatenation (28.21 vs 27.97 mIoU).
  • The vertical-slice fusion module costs about 0.6 GB of extra validation GPU memory for a 1.14-point mIoU gain.

Reading between the lines

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

  • The hand-picked slice boundaries are a dataset-specific choice; a learnable slicing mechanism could preserve the gains without retuning per dataset.
  • The height-aware channel weighting is not tied to occupancy prediction and could transfer to other voxel-based 3D perception tasks, such as LiDAR-camera 3D detection or scene completion.
  • The reported weakness on flat-ground classes suggests height alone cannot separate semantically different but geometrically similar ground classes; injecting LiDAR intensity or surface orientation would be a natural next test.
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

4 major / 6 minor

Summary. The paper proposes SliceSemOcc, a multimodal (camera + LiDAR) 3D semantic occupancy prediction framework. The core ideas are (i) a Vertical Slice Fusion module that extracts features from global and local height slices of the voxel volume, (ii) a per-height channel-attention module (SEAttention3D) that preserves height resolution during pooling, and (iii) a global–local cross-attention fusion. Experiments on nuScenes-SurroundOcc and nuScenes-OpenOccupancy report mIoU of 28.2 and 22.9, respectively, with ablations claiming that each component contributes positively. The paper also reports a memory overhead of 0.6 GB for the full module.

Significance. If the reported internal improvements are robust, the work provides a useful, incremental architectural contribution: height-aware channel attention and dual-scale vertical slicing are intuitive and align with observed height distributions of small objects. The component-wise ablations (Tables 3–4, Section 4.3) are a strength, as is the explicit reporting of memory cost and the candid discussion of degraded flat-ground categories. However, the headline comparison is confounded by an uncontrolled baseline, the slice boundaries are derived from the same dataset on which the method is evaluated, and the second dataset result ties an existing method (OccLoff) without acknowledgement. No code, training hyperparameters, or multiple-seed statistics are provided, making the small reported gains difficult to verify.

major comments (4)
  1. [§4.2, Tables 1 and 3] The central claim of a 14.2% relative mIoU gain over M-CONet (24.7 → 28.2) is confounded by an uncontrolled baseline. Table 3 shows that the authors' own no-slice baseline achieves 26.98 mIoU, which is 2.28 points above the published M-CONet number. Section 4.1 lists many implementation differences from M-CONet (InverseMatrixVT3D projection, FPN, VoxelNet, focal+Lovasz+affinity losses), so the gap between 26.98 and 24.7 likely reflects training recipe and view-transformation differences, not vertical slicing. The internal contribution of the proposed modules is therefore 26.98 → 28.21, i.e., +1.23 mIoU (~4.6% relative), not 14.2%. The small-object gains in Section 4.2 are also computed relative to M-CONet's published class-wise numbers and inherit the same confound. The paper must either train M-CONet under its own pipeline or present the internal ablation as the primary result.
  2. [§3.4, Fig. 1, Table 4] The local slice boundaries [-5,-3], [-3,-2], [-2,-1], [-1,0], [0,1], [1,3] m are explicitly chosen based on the object height distribution of nuScenes-SurroundOcc (Fig. 1) and then evaluated on the validation set of the same dataset. The paper never demonstrates that these boundaries were fixed using only the training split, nor does it test transfer to another dataset. Table 4 shows that the boundary choice matters (28.21 vs. 27.67/27.75), so this is not a negligible design detail. The authors should state whether the boundaries were computed from training statistics only, and provide sensitivity/transfer experiments (e.g., applying the same boundaries to nuScenes-OpenOccupancy or reporting dataset-specific boundaries).
  3. [§4.2, Table 2] On nuScenes-OpenOccupancy, SliceSemOcc reports 22.9 mIoU, which is identical to the 22.9 mIoU of OccLoff [44] listed in the same table. The text says the method 'improves mIoU from 20.1 (M-CONet) to 22.9' and claims 'substantial gains', but it does not acknowledge that this is a tie with the current state-of-the-art OccLoff. This is a fairness and completeness issue: the reader cannot assess whether the proposed method advances the state of the art when the leading comparison is omitted from the narrative. Please revise the discussion to state the tie explicitly and, if possible, add a statistical comparison (multiple seeds) to distinguish the methods.
  4. [§4.1, §4.3] The empirical claims rest on small differences (e.g., 27.97 vs. 28.21 for fusion strategy; 28.08 vs. 28.21 for SEAttention3D), but the paper reports no multiple-seed runs, error bars, or significance tests. Training hyperparameters (optimizer, learning rate, batch size, number of epochs, voxel size details) are not provided, and no code is released. For a deep-learning paper in a field where single-run differences of 0.1–0.2 mIoU are common, this lack of uncertainty quantification is load-bearing. At minimum, the authors should report the number of runs, mean and std, and provide the training configuration in an appendix.
minor comments (6)
  1. [Table 4] Typo: 'moudle' should be 'module' in the caption.
  2. [Fig. 3] The caption appears to contain duplicated 'Inception' text; Fig. 3 shows SEAttention3D, not inception modules. Please clean up the figure/caption.
  3. [Fig. 1] The height distribution plot has no axis labels or units. Without these, the justification for the chosen slice boundaries cannot be assessed.
  4. [Table 2 title] The title says 'nuScenes-Occupancy', but the text and the rest of the paper refer to 'nuScenes-OpenOccupancy'. Please unify the name.
  5. [Eq. (5)] After reducing channels to one with a 3D convolution, A_global and A_local have shape B×1×X×Y×Z; the elementwise multiplication with F_global (B×C×X×Y×Z) requires broadcasting. Please state this explicitly.
  6. [§3.1] The problem statement says labels are in [0,16], but the evaluation uses 16 semantic classes plus empty. Please clarify whether 0 is empty and classes are 1–16, and whether '16' is inclusive.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: reported gains are empirical mIoU values; the hand-chosen slice boundaries are a dataset-informed design choice, not a fitted parameter renamed as prediction.

full rationale

The paper's central claims are empirical: SliceSemOcc achieves 28.2 mIoU on nuScenes-SurroundOcc and 22.9 on nuScenes-OpenOccupancy, with ablations isolating the contributions of global/local slices, SEAttention3D, and cross-attention fusion. No equation defines the output in terms of the input, no fitted parameter is renamed as a prediction, and no load-bearing self-citations or imported uniqueness theorems appear. The local-slice intervals ([-5,-3], [-3,-2], [-2,-1], [-1,0], [0,1], [1,3]) are chosen from the object height distribution of nuScenes-SurroundOcc (Section 3.4, Fig. 1) and then evaluated on the same dataset's validation split; Table 4 selects the best-performing partitioning on that validation set. This is a legitimate experimental concern (validation-set-informed architecture selection / multiple testing), but it is not circular in the derivational sense: the reported mIoU is a measured quantity, not constructed from the height-distribution statistics. Similarly, the baseline mismatch with M-CONet (Table 3's no-slice baseline reaches 26.98 mIoU vs. M-CONet's 24.7, with other pipeline differences listed in Section 4.1) undermines the headline comparison as a controlled attribution, but this is an experimental-control issue, not a circular-reasoning issue. Under the stated criteria requiring a specific reduction by construction or a fitted parameter called a prediction, no circular step can be identified.

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

The central claim rests on hand-chosen architectural hyperparameters (slice boundaries, reduction ratio, loss weights) and on trust in the benchmark labels, cited backbone components, and baseline numbers. No new physical entities are introduced. The slice-boundary choice is the most dataset-specific element and the one most likely to limit generalization.

free parameters (4)
  • Local slice boundaries for height binning = [-5,-3],[-3,-2],[-2,-1],[-1,0],[0,1],[1,3] meters
    Chosen from the class height distribution of nuScenes-SurroundOcc (Section 3.4, Fig. 1) and evaluated on the same dataset; a dataset-specific hand-tuned design choice.
  • SEAttention3D reduction ratio r = Not reported in the paper
    Architecture hyperparameter used in the excitation step (Section 3.3); the value affects capacity and the reported mIoU, but is never stated.
  • Number and width of local slices = 6 intervals as listed above
    Implied by the binning; the ablation in Table 4 shows this scheme beats alternatives on the same dataset, but the search over schemes uses the same evaluation split.
  • Loss weights in Ltotal = Implicitly all 1.0 in Eq. 6
    The total loss sums focal, Lovasz, and two affinity losses without stated weighting; a hand-set choice that influences final mIoU.
assumptions (5)
  • domain assumption Ground-truth occupancy labels from OpenOccupancy and SurroundOcc are correct and complete enough to measure mIoU.
    The paper trusts the benchmark labels without analysis of label noise (Section 4.1).
  • domain assumption The projection method of InverseMatrixVT3D [24] and the VoxelNet [46] backbone produce well-aligned camera and LiDAR voxel features.
    The whole fusion pipeline relies on these components, which are cited but not re-verified (Section 3.2).
  • ad hoc to paper The aggregate height distribution in Fig. 1 is representative of both training and validation scenes.
    Used to justify the specific local slice partitions; no split information is given (Section 3.4).
  • standard math Focal, Lovasz, and affinity losses are correctly implemented and combined.
    Loss definitions are taken from prior works and summed in Eq. 6 without derivation.
  • domain assumption Baseline numbers in Tables 1 and 2 are accurately reproduced from the cited sources and are comparable in training settings.
    No code is provided to reproduce baselines or the proposed method, so comparability is assumed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SliceSemOcc: Vertical Slice Based Multimodal 3D Semantic Occupancy Representation." pith.science (2026). https://pith.science/paper/OM5OAWUW

@misc{pith2026250903999,
  author       = {Pith},
  title        = {Pith review of: SliceSemOcc: Vertical Slice Based Multimodal 3D Semantic Occupancy Representation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OM5OAWUW}},
  note         = {Machine review of arXiv:2509.03999}
}
read the original abstract

Driven by autonomous driving's demands for precise 3D perception, 3D semantic occupancy prediction has become a pivotal research topic. Unlike bird's-eye-view (BEV) methods, which restrict scene representation to a 2D plane, occupancy prediction leverages a complete 3D voxel grid to model spatial structures in all dimensions, thereby capturing semantic variations along the vertical axis. However, most existing approaches overlook height-axis information when processing voxel features. And conventional SENet-style channel attention assigns uniform weight across all height layers, limiting their ability to emphasize features at different heights. To address these limitations, we propose SliceSemOcc, a novel vertical slice based multimodal framework for 3D semantic occupancy representation. Specifically, we extract voxel features along the height-axis using both global and local vertical slices. Then, a global local fusion module adaptively reconciles fine-grained spatial details with holistic contextual information. Furthermore, we propose the SEAttention3D module, which preserves height-wise resolution through average pooling and assigns dynamic channel attention weights to each height layer. Extensive experiments on nuScenes-SurroundOcc and nuScenes-OpenOccupancy datasets verify that our method significantly enhances mean IoU, achieving especially pronounced gains on most small-object categories. Detailed ablation studies further validate the effectiveness of the proposed SliceSemOcc framework.

Figures

Figures reproduced from arXiv: 2509.03999 by the authors.

Figure 1
Figure 1. Height distribution of different classes on nuScenes-SurroundOcc. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Overall architecture of SliceSemOcc. featuresFL ∈ R B×C×X×Y ×Z.Both FC and FL pass through the Vertical Slice Fu￾sion module which divides each along the height-axis into global slices covering the entire range and local slices focusing on key height bands. Each slice is first processed by our SEAttention3D module. Then, a bidirectional cross attention mechanism fuses global and local slice features through mutual r… view at source ↗
Figure 3
Figure 3. Detailed architecture of the SEAttention3D module. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Detailed architecture of (a) Vertical Slice Fusion (VSF) module and (b) [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Qualitative performance on nuScenes-SurroundOcc dataset. [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 19 canonical work pages

  1. [44]

    In: W ACV

    Zhang, J., Ding, Y., Liu, Z.: Occloff: Learning optimized feature fu- sion for 3d occupancy prediction. In: W ACV. pp. 3096–3106 (2025). https://doi.org/10.1109/W ACV61041.2025.00306

    arXiv 2025

  2. [2]

    In: CVPR

    Berman, M., Triki, A.R., Blaschko, M.B.: The lovasz-softmax loss: A tractable surrogate for the optimization of the intersection-over- union measure in neural networks. In: CVPR. pp. 4413–4421 (2018). https://doi.org/10.1109/CVPR.2018.00464

    arXiv 2018

  3. [4]

    In: CVPR

    Cao, A., de Charette, R.: Monoscene: Monocular 3d semantic scene completion. In: CVPR. pp. 3981–3991 (2022). https://doi.org/10.1109/CVPR52688.2022.00396

    arXiv 2022

  4. [5]

    In: CVPR

    Chen, X., Lin, K.Y., Qian, C., Zeng, G., Li, H.: 3d sketch-aware semantic scene completion via semi-supervised structure prior. In: CVPR. pp. 4193–4202 (2020)

    work page 2020

  5. [6]

    In: CVPR

    Chen, X., Ma, H., Wan, J., Li, B., Xia, T.: Multi-view 3d object detection network for autonomous driving. In: CVPR. pp. 1907–1915 (2017)

    work page 1907

  6. [7]

    In: CVPR

    Chen, X., Zhang, T., Wang, Y., Wang, Y., Zhao, H.: Futr3d: A unified sensor fusion framework for 3d detection. In: CVPR. pp. 172–181 (2023)

    work page 2023

  7. [8]

    In: CVPR

    Chi, X., Liu, J., Lu, M., Zhang, R., Wang, Z., Guo, Y., Zhang, S.: Bev-san: Accurate bev 3d object detection via slice attention networks. In: CVPR. pp. 17461–17470 (Jun 2023). https://doi.org/10.1109/CVPR52729.2023.01675

    arXiv 2023

  8. [9]

    In: CVPR

    He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: CVPR. pp. 770–778 (2016). https://doi.org/10.1109/CVPR.2016.90

    doi:10.1109/cvpr.2016.90 2016

Show all 43 references
  1. [10]

    CoRR abs/2203.17054 (2022)

    Huang, J., Huang, G.: Bevdet4d: Exploit temporal cues in multi- camera 3d object detection. CoRR abs/2203.17054 (2022). https://doi.org/10.48550/ARXIV.2203.17054

    2022 doi arXiv

  2. [11]

    aeXiv preprint arXiv:2112.11790 (2021)

    Huang, J., Huang, G., Zhu, Z., Du, D.: Bevdet: High-performance multi-camera 3d object detection in bird-eye-view. aeXiv preprint arXiv:2112.11790 (2021)

    2021 arXiv

  3. [12]

    In: CVPR

    Huang, Y., Zheng, W., Zhang, Y., Zhou, J., Lu, J.: Tri-perspective view for vision- based 3d semantic occupancy prediction. In: CVPR. pp. 9223–9232 (2023)

    2023

  4. [13]

    In: AAAI (2023)

    Jiang, Y., Zhang, L., Miao, Z., Zhu, X., Gao, J., Hu, W., Jiang, Y.G.: Polar- former: multi-camera 3d object detection with polar transformer. In: AAAI (2023). https://doi.org/10.1609/aaai.v37i1.25185

    2023 doi

  5. [14]

    In: CVPR (2023) Vertical Slice–Based Multimodal 3D Semantic Occupancy Representation 13

    Khurana, T., Hu, P., Held, D., Ramanan, D.: Point cloud forecasting as a proxy for 4d occupancy forecasting. In: CVPR (2023) Vertical Slice–Based Multimodal 3D Semantic Occupancy Representation 13

    2023

  6. [15]

    In: CVPR

    Li, J., Han, K., Wang, P., Liu, Y., Yuan, X.: Anisotropic convolutional net- works for 3d semantic scene completion. In: CVPR. pp. 3348–3356 (2020). https://doi.org/10.1109/CVPR42600.2020.00341

    2020

  7. [16]

    In: AAAI

    Li, Y., Ge, Z., Yu, G., Yang, J., Wang, Z., Shi, Y., Sun, J., Li, Z.: Bevdepth: Acquisition of reliable depth for multi-view 3d object detection. In: AAAI. vol. 37, pp. 1477–1485 (2023)

    2023

  8. [17]

    In: ECCV

    Li, Z., Wang, W., Li, H., Xie, E., Sima, C., Lu, T., Yu, Q., Dai, J.: Bevformer: Learning bird’s-eye-view representation from multi-camera images via spatiotem- poral transformers. In: ECCV. pp. 1–18 (2022). https://doi.org/10.1007/978-3- 031-20077-9 1

    2022 doi

  9. [18]

    In: ECCV

    Liang, M., Yang, B., Wang, S., Urtasun, R.: Deep continuous fusion for multi-sensor 3d object detection. In: ECCV. pp. 641–656 (2018)

    2018

  10. [19]

    NeurIPS 35, 10421–10434 (2022)

    Liang, T., Xie, H., Yu, K., Xia, Z., Lin, Z., Wang, Y., Tang, T., Wang, B., Tang, Z.: Bevfusion: A simple and robust lidar-camera fusion framework. NeurIPS 35, 10421–10434 (2022)

    2022

  11. [20]

    In: CVPR

    Lin, T., Doll´ ar, P., Girshick, R.B., He, K., Hariharan, B., Belongie, S.J.: Feature pyramid networks for object detection. In: CVPR. pp. 936–944. IEEE Computer Society (2017). https://doi.org/10.1109/CVPR.2017.106

    2017 doi

  12. [21]

    In: ICCV

    Lin, T., Goyal, P., Girshick, R.B., He, K., Doll´ ar, P.: Focal loss for dense object detection. In: ICCV. pp. 2999–3007. IEEE Computer Society (2017). https://doi.org/10.1109/ICCV.2017.324

    2017 doi

  13. [22]

    arXiv preprint arXiv:2310.11239 (2023)

    Liu, X., Gong, M., Fang, Q., Xie, H., Li, Y., Zhao, H., Feng, C.: Lidar-based 4d occupancy completion and forecasting. arXiv preprint arXiv:2310.11239 (2023)

    2023 arXiv

  14. [23]

    In: ICRA

    Liu, Z., Tang, H., Amini, A., Yang, X., Mao, H., Rus, D.L., Han, S.: Bevfu- sion: Multi-task multi-sensor fusion with unified bird’s-eye view representation. In: ICRA. pp. 2774–2781 (2023). https://doi.org/10.1109/ICRA48891.2023.10160968

    2023

  15. [24]

    arXiv preprint arXiv:2401.12422 (2024)

    Ming, Z., Berrio, J.S., Shan, M., Worrall, S.: Inversematrixvt3d: An efficient projection matrix-based approach for 3d occupancy prediction. arXiv preprint arXiv:2401.12422 (2024)

    2024 arXiv

  16. [25]

    Ming, Z., Berrio, J.S., Shan, M., Worrall, S.: Occfusion: Multi-sensor fu- sion framework for 3d semantic occupancy prediction. TIV pp. 1–13 (2024). https://doi.org/10.1109/TIV.2024.3453293

    2024

  17. [26]

    In: ICCV

    Nabati, R., Qi, H.: Centerfusion: Center-based radar and camera fusion for 3d object detection. In: ICCV. pp. 1527–1536 (2021)

    2021

  18. [27]

    IEEE Robotics Autom

    Pan, J., Wang, Z., Wang, L.: Co-occ: Coupling explicit feature fusion with volume rendering regularization for multi-modal 3d semantic occu- pancy prediction. IEEE Robotics Autom. Lett. 9(6), 5687–5694 (2024). https://doi.org/10.1109/LRA.2024.3396092

    2024

  19. [28]

    CoRR abs/2309.09502 (2023)

    Pan, M., Liu, J., Zhang, R., Huang, P., Li, X., Liu, L., Zhang, S.: Renderocc: Vision-centric 3d occupancy prediction with 2d rendering supervision. CoRR abs/2309.09502 (2023). https://doi.org/10.48550/ARXIV.2309.09502

    2023 doi arXiv

  20. [29]

    In: ECCV

    Philion, J., Fidler, S.: Lift, splat, shoot: Encoding images from arbitrary cam- era rigs by implicitly unprojecting to 3d. In: ECCV. pp. 194–210 (2020). https://doi.org/10.1007/978-3-030-58568-6 12

    2020 doi

  21. [30]

    In: ICCV

    Qi, C.R., Litany, O., He, K., Guibas, L.J.: Deep hough voting for 3d object detec- tion in point clouds. In: ICCV. pp. 9277–9286 (2019)

    2019

  22. [31]

    In: CVPR

    Qi, C.R., Liu, W., Wu, C., Su, H., Guibas, L.J.: Frustum pointnets for 3d object detection from rgb-d data. In: CVPR. pp. 918–927 (2018)

    2018

  23. [32]

    Rold˜ ao, L., de Charette, R., Verroust-Blondet, A.: Lmscnet: Lightweight multiscale 3d semantic completion. In: 3DV. pp. 111–119 (2020). https://doi.org/10.1109/3DV50981.2020.00021 14 H. Huang, H. Sun et al

    2020

  24. [33]

    In: CVPR

    Shi, S., Guo, C., Jiang, L., Wang, Z., Shi, J., Wang, X., Li, H.: Pv-rcnn: Point-voxel feature set abstraction for 3d object detection. In: CVPR. pp. 10529–10538 (2020)

    2020

  25. [34]

    In: CVPR

    Wang, C., Ma, C., Zhu, M., Yang, X.: Pointaugmenting: Cross-modal aug- mentation for 3d object detection. In: CVPR. pp. 11789–11798 (2021). https://doi.org/10.1109/CVPR46437.2021.01162

    2021

  26. [36]

    In: Conference on Robot Learning

    Wang, Y., Guizilini, V.C., Zhang, T., Wang, Y., Zhao, H., Solomon, J.: Detr3d: 3d object detection from multi-view images via 3d-to-2d queries. In: Conference on Robot Learning. pp. 180–191. PMLR (2022)

    2022

  27. [37]

    In: IROS

    Wang, Z., Jia, K.: Frustum convnet: Sliding frustums to aggregate local point- wise features for amodal 3d object detection. In: IROS. pp. 1742–1749 (2019). https://doi.org/10.1109/IROS40897.2019.8968513

    2019

  28. [38]

    In: ICCV

    Wei, Y., Zhao, L., Zheng, W., Zhu, Z., Zhou, J., Lu, J.: Surroundocc: Multi- camera 3d occupancy prediction for autonomous driving. In: ICCV. pp. 21672– 21683 (2023). https://doi.org/10.1109/ICCV51070.2023.01986

    2023

  29. [39]

    In: AAAI

    Yan, X., Gao, J., Li, J., Zhang, R., Li, Z., Huang, R., Cui, S.: Sparse single sweep lidar point cloud segmentation via learning contextual shape priors from scene completion. In: AAAI. vol. 35, pp. 3101–3109 (2021)

    2021

  30. [40]

    Sensors 18(10) (2018)

    Yan, Y., Mao, Y., Li, B.: Second: Sparsely embedded convolutional detection. Sensors 18(10) (2018). https://doi.org/10.3390/S18103337

    2018 doi

  31. [41]

    In: CVPR

    Yang, Z., Sun, Y., Liu, S., Jia, J.: 3dssd: Point-based 3d single stage object detector. In: CVPR. pp. 11040–11048 (2020)

    2020

  32. [42]

    arXiv preprint arXiv:1812.05276 (2018)

    Yang, Z., Sun, Y., Liu, S., Shen, X., Jia, J.: Ipod: Intensive point-based object detector for point cloud. arXiv preprint arXiv:1812.05276 (2018)

    2018 arXiv

  33. [43]

    arXiv preprint arXiv:2111.08318 (2021)

    Ye, M., Wan, R., Xu, S., Cao, T., Chen, Q.: Drinet++: Efficient voxel-as-point point cloud segmentation. arXiv preprint arXiv:2111.08318 (2021)

    2021 arXiv

  34. [45]

    arXiv preprint arXiv:2008.01550 (2020)

    Zhou, H., Zhu, X., Song, X., Ma, Y., Wang, Z., Li, H., Lin, D.: Cylinder3d: An ef- fective 3d framework for driving-scene lidar semantic segmentation. arXiv preprint arXiv:2008.01550 (2020)

    2008 arXiv

  35. [46]

    In: CVPR

    Zhou, Y., Tuzel, O.: Voxelnet: End-to-end learning for point cloud based 3d object detection. In: CVPR. pp. 4490–4499 (2018). https://doi.org/10.1109/CVPR.2018.00472

    2018

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.