Pith. sign in

REVIEW 4 major objections 6 minor 49 references

LDRFusion: A LiDAR-Dominant multimodal refinement framework for 3D object detection

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read LDRFusion claims a LiDAR-dominant two-stage refinement scheme improves 3D detection by adding camera pseudo points only in the second stage, reporting 85.94 mAP on KITTI.

desk verdict A LiDAR-dominant cascade with a plausible HPR module, but the gains are small, unreplicated, and under-specified, so it deserves conditional review rather than acceptance as-is. read the letter →

arxiv 2507.16224 v2 pith:CBASHDXT submitted 2025-07-22 cs.CV

Jijun Wang , Yan Wu , Yujian Mo , Junqiao Zhao , Jun Yan , Yinghao Hu This is my paper · ORCID
classification cs.CV
keywords LiDAR-camerafusion3DobjectdetectionpseudopointclouddepthcompletioncascaderefinementresidualencodingKITTIbenchmarkautonomousdriving
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

LDRFusion is a LiDAR-dominant two-stage refinement framework for 3D object detection. It first generates proposals from LiDAR alone, then refines them with dense pseudo point clouds produced by depth completion, and finally merges the two stages' instance-level predictions. The paper reports that this asymmetric cascade outperforms the SFD baseline on KITTI at every difficulty, with the largest gain on hard cases, and sets a multi-class validation mAP of 78.71. The point is that noisy pseudo points should be introduced only at a refinement stage, where their density helps, rather than fused symmetrically from the start. If correct, the design offers a computationally cheap way to upgrade pseudo-point detectors without retraining from scratch.

What carries the argument

The load-bearing object is the hierarchical pseudo point residual encoding (HPR) module. For each pseudo point in a cropped region of interest, HPR finds K neighboring pseudo points by searching nearby image pixels, then iteratively updates each point's feature by concatenating its feature residual relative to the centroid with its own feature and reweighting the result by an MLP applied to the positional residual; aggregation uses an MLP rather than max pooling because RGB features break rotational invariance. This module supplies the second stage's pseudo-point features, while the cascade itself -- LiDAR-only first stage, fused second stage, and weighted instance-level merging $D = \alpha D^L + (1-\alpha) D^M$ with $\alpha = 0.5$ -- is the architectural mechanism the paper credits for the gains.

What would settle it

Repeat the KITTI car experiment with the same training configuration for both SFD and LDRFusion across ten random seeds; if the 95% confidence interval for the moderate-difficulty AP difference includes zero, the central claim of consistent improvement is not supported.

Watch

Extended reading notes

Core claim

The central claim is that detection accuracy improves when the fusion is ordered rather than symmetric: a LiDAR-only stage first produces well-localized proposals, a second stage fuses those proposals with pseudo point clouds to catch hard instances, and the two sets of instance-level outputs are combined with a fixed weight (alpha equal to 0.5). The paper reports KITTI test 3D car AP (R40) of 91.92/85.47/80.43 for easy/moderate/hard (mAP 85.94), compared with SFD's 91.73/84.76/77.92, and a multi-class KITTI validation mAP of 78.71, a 0.92-point improvement over the prior best. It also reports nuScenes mAP/NDS of 67.8/71.0, surpassing the two-stage MVP baseline, and an inference speed of 10.0 FPS, only 0.2 FPS below SFD. The paper argues these results show LiDAR should remain the localization authority while camera-derived pseudo points act as a second-stage enhancer.

Load-bearing premise

The load-bearing premise is that the reported gains come from the proposed design and not from run-to-run randomness or hidden tuning, since the paper reports no error bars and leaves some HPR hyperparameters (K and the search radius n) unspecified.

Editorial extensions

If this is right

  • On the KITTI test set, the framework reports 3D car AP (R40) of 91.92/85.47/80.43 at easy/moderate/hard, an mAP of 85.94, beating the SFD baseline by 0.19, 0.71, and 2.51 points.
  • Because the largest improvement is on hard examples, the design specifically helps where sparse LiDAR is weakest.
  • The nuScenes result (mAP 67.8, NDS 71.0, surpassing the two-stage MVP baseline) indicates the refinement can be attached to other pseudo-point detectors.
  • Keeping pseudo points out of the first stage avoids injecting depth-completion noise into proposal localization, while instance-level fusion keeps inference near 10 FPS, only 0.2 FPS below SFD.
  • On joint multi-class validation, the method reaches 78.71 mAP, a 0.92-point gain over the previous best, but the improvement concentrates on cars rather than pedestrians and cyclists.

Reading between the lines

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

  • If the reported KITTI gains hold under repeated-seed evaluation, the same LiDAR-dominant cascade should transfer to other pseudo-point detectors, not just depth-completion ones, because the design only assumes a noisy dense auxiliary cloud.
  • The class asymmetry (largest gains on cars, second-best on pedestrians and cyclists) suggests the refinement helps most where depth estimation is reliable; a testable extension is to gate the second stage by estimated pseudo-point confidence per class.
  • Unreported HPR neighborhood parameters (K and the search radius n) mean the method's sensitivity to those choices is unknown; measuring performance across K and n would clarify how much of the gain is architectural versus tuned.
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 LDRFusion, a two-stage LiDAR-dominant fusion framework for 3D object detection. In the first stage, only raw LiDAR point clouds are used to generate proposals; in the second stage, pseudo point clouds obtained from depth completion are incorporated, and the instance-level outputs of both stages are combined with a weighted fusion at inference. A hierarchical pseudo point residual encoding (HPR) module is introduced to capture local structure in pseudo point clouds. Experiments on KITTI and nuScenes report consistent improvements over the SFD baseline, a multi-class validation mAP of 78.71, and comparable inference speed. Ablations in Tables 6 and 7 support the design choices, but the empirical case is weakened by the absence of error bars, unstated HPR hyperparameters, and an ambiguously defined fusion rule.

Significance. If the reported results are reproducible, the paper makes a useful incremental contribution: a simple, efficient LiDAR-dominant refinement strategy that can be plugged into pseudo-point-based detectors, plus an HPR module that improves local representation. The architecture is coherent and the ablations are informative. However, the central empirical claim rests on single-run AP differences that are mostly below 1 mAP on the KITTI validation set, and several reproducibility-critical details are missing. The paper would be a solid systems-style contribution once the statistical reliability of the gains and the missing implementation details are addressed.

major comments (4)
  1. [§3.3 (HPR module)] The HPR neighborhood size K and pixel search range n are introduced in §3.3 but their numerical values are never given, neither in the method section nor in §4.2. Since HPR is one of the two main contributions and Table 6 credits it with a 0.26 mAP gain, the unstated parameters make the module irreproducible and prevent an assessment of sensitivity. Please specify K and n, and ideally include a sensitivity study over these values.
  2. [§3.2, Eq. (10)] The inference-time fusion rule D = αDL + (1−α)DM is not well defined: DL and DM are sets of bounding-box instances, not numeric vectors, so the weighted sum requires a matching or association rule and a score-merging rule (e.g., IoU matching followed by score weighting, then NMS). The paper does not state how the two instance sets are paired, how duplicate detections are removed, or whether α=0.5 is applied per matched pair. This is central to the claimed LiDAR-dominant refinement and must be specified precisely.
  3. [§4.4, Tables 6 and 7] All reported AP values are single runs without error bars or repeated-seed statistics. The headline ablation gains in Table 6 are +0.53/+0.50/+0.73 mAP over the SFD baseline, and the HPR-only gain is +0.26 mAP; on the 3769-frame KITTI validation split such differences are typically within run-to-run training variance. The paper should provide multiple seeds (mean ± std) or a paired statistical test to support the central claim that the architectural components, not noise or training-configuration variation, cause the improvements.
  4. [Tables 6 and 7] The ablation baselines appear inconsistent. Table 6 (a) reports the SFD baseline as 95.47/88.56/85.74, while Table 7 (II), which uses real+pseudo point clouds in both stages with instance-level fusion, reports 95.26/88.26/85.73. If (II) is meant to be the SFD-style configuration, the two numbers should coincide; if not, the Table 7 comparison lacks a stated baseline. Please clarify the exact configuration of Table 7 (II) and how it relates to Table 6 (a), since the LiDAR-dominant conclusion depends on this comparison.
minor comments (6)
  1. [Throughout] Typographical errors include 'soley' in the abstract, 'genrated' in §1, 'fisrt' in §3.2, and 'strucure' in the Figure 3 caption; please proofread the text.
  2. [§3.3, Eq. (13)] The refinement loop in Eq. (13) concatenates s1 through st+1 but does not state the total number of iterations T or the initial condition for t; please define the loop bounds explicitly.
  3. [§3.4, Eq. (14)] The loss symbols LL_aux and LC_aux are confusing because the subscripts do not match the modalities they supervise (pseudo point clouds vs. raw point clouds). Please rename or define these terms explicitly.
  4. [Table 4] The row 'MVP + 2 stage + LPRFusion' presumably should read 'LDRFusion'; the text describes LDRFusion, so the table label appears to be a typo.
  5. [Table 5] The inference-speed comparison does not state measurement conditions such as hardware, batch size, input resolution, or whether depth completion time is included; please provide these details.
  6. [§3.2.1, Eq. (2)] Part-Sensitive Warping (PSW) is used in Eq. (2) but is never defined and no precise reference is given for it; please define the operation or cite the source.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's claims are empirical and benchmarked against external datasets and independent baselines.

full rationale

LDRFusion is an empirical systems paper rather than a derivation. The central claims are performance numbers on the KITTI test/validation sets and the nuScenes validation set, measured against external baselines such as SFD, Voxel R-CNN, CasA, and MVP. The proposed LDRFusion components are constructed from explicit equations (Eqs. 1-14) describing the cascade refinement, HPR residual encoding, fusion, and loss; none of these equations is defined in terms of the result it is used to predict. The hyperparameters alpha=0.5 and the loss weights lambda1=1.0, lambda2=0.5, lambda3=1.0 are fixed design choices, not fitted to the test set, so no quantity labeled a prediction reduces to a fitted input by construction. The second stage consumes first-stage proposals, but that is an architectural dependency, not a circular definition. Self-citations exist (e.g., SQD [22], a prior survey [1], and an arXiv preprint [13]), but they are used only as related work or motivational context and are not load-bearing for the paper's empirical conclusions. The absence of error bars, code, and stated HPR neighbor-search parameters is a reproducibility concern, not a circularity concern: the reported gains are externally falsifiable on public benchmarks regardless of whether they exceed run-to-run variance. Therefore no specific circular step can be exhibited, and the appropriate finding is no significant circularity.

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

The paper is an empirical engineering contribution; its axioms are about experimental setup and protocol trust rather than mathematical grounding. The free parameters are hyperparameters that influence the reported numbers but are not derived from first principles.

free parameters (4)
  • Fusion weight alpha = 0.5
    Equation 10 averages first-stage and second-stage instance predictions; the weight is hand-set.
  • Loss weights lambda_1, lambda_2, lambda_3 = 1.0, 0.5, 1.0
    Equation 14 balances RPN, refinement, auxiliary, and fusion losses; values are chosen without sensitivity analysis.
  • HPR neighborhood size K
    Section 3.3 defines a K-neighbor search but never states the value of K used in experiments.
  • HPR pixel search range n
    Section 3.3 sets the neighborhood via pixel offsets +-n, but n is not specified.
assumptions (3)
  • domain assumption The fixed pre-trained TWISE depth completion network yields pseudo point clouds reliable enough for refinement.
    Section 4.2 introduces TWISE without ablation on depth quality; the second stage depends entirely on these pseudo points.
  • domain assumption KITTI and nuScenes evaluation protocols, including the train/validation partition and AP computation, are applied correctly.
    Section 4.1 states standard protocols but no code is released, so correctness cannot be checked.
  • ad hoc to paper The two-stage cascade with weighted instance-level fusion is superior to symmetric fusion.
    Tables 6-7 support this on one dataset with small margins and no statistical test; it is presented as a design conclusion rather than a derived result.

how reviews work

0 comments
Cite this review

Pith. "Pith review of LDRFusion: A LiDAR-Dominant multimodal refinement framework for 3D object detection." pith.science (2026). https://pith.science/paper/CBASHDXT

@misc{pith2026250716224,
  author       = {Pith},
  title        = {Pith review of: LDRFusion: A LiDAR-Dominant multimodal refinement framework for 3D object detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CBASHDXT}},
  note         = {Machine review of arXiv:2507.16224}
}
read the original abstract

Existing LiDAR-Camera fusion methods have achieved strong results in 3D object detection. To address the sparsity of point clouds, previous approaches typically construct spatial pseudo point clouds via depth completion as auxiliary input and adopts a proposal-refinement framework to generate detection results. However, introducing pseudo points inevitably brings noise, potentially resulting in inaccurate predictions. Considering the differing roles and reliability levels of each modality, we propose LDRFusion, a novel Lidar-dominant two-stage refinement framework for multi-sensor fusion. The first stage soley relies on LiDAR to produce accurately localized proposals, followed by a second stage where pseudo point clouds are incorporated to detect challenging instances. The instance-level results from both stages are subsequently merged. To further enhance the representation of local structures in pseudo point clouds, we present a hierarchical pseudo point residual encoding module, which encodes neighborhood sets using both feature and positional residuals. Experiments on the KITTI dataset demonstrate that our framework consistently achieves strong performance across multiple categories and difficulty levels.

Figures

Figures reproduced from arXiv: 2507.16224 by the authors.

Figure 1
Figure 1. Visualization of detection results from different models: (a) using real point clouds, (b) using real and pseudo point clouds, and (c) using the proposed refinement strategy (ours). Ground truths are shown in red boxes, predictions in green. Black points inside predictions represent pseudo point clouds, while red points indicate real ones [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Comparison of different fusion paradigms. Previous approaches rely on a symmetric fusion strategy. In contrast, we adopt a LiDAR-dominant two-stage refinement scheme, which integrates instance-level outputs from multiple stages. and feature residuals of pseudo point clouds. This design cap￾tures local structural and contextual relationships while main￾taining computational efficiency. Our contributions can be summar… view at source ↗
Figure 3
Figure 3. Overall strucure of LDRFusion. Our proposed framework primarily focuses on box refinement through a two-stage optimization process. In the first stage, the original point clouds are exclusively utilized, while the second stage integrates dense semantic information from pseudo point clouds to generate precise detection results. When incorporating auxiliary pseudo point clouds, we employ HPR to effectively extract loc… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 46 canonical work pages

  1. [1]

    Review the state-of-the-art technologies of semantic segmentation based on deep learning

    Yujian Mo, Yan Wu, Xinneng Yang, Feilin Liu, and Yujun Liao. Review the state-of-the-art technologies of semantic segmentation based on deep learning. Neurocomputing, 493:626–646, 2022. ISSN 0925-2312. doi: https: //doi.org/ 10.1016/j.neucom.2022.01.005

    doi:10.1016/j.neucom.2022.01.005 2022

  2. [2]

    3d object detection for autonomous driving: A survey

    Rui Qian, Xin Lai, and Xirong Li. 3d object detection for autonomous driving: A survey. Pattern Recognition, 130: 108796, 2022

    work page 2022

  3. [3]

    3d object detection for autonomous driving: A comprehensive survey

    Jiageng Mao, Shaoshuai Shi, Xiaogang Wang, and Hong- sheng Li. 3d object detection for autonomous driving: A comprehensive survey. International Journal of Computer Vision, 131(8):1909–1963, 2023. Preprint – LDRFusion: A LiDAR-Dominant multimodal refinement framework for 3D object detection 6 Table 1: Performance for car detection on the KITTI test set (%...

    work page 1909

  4. [4]

    Adversarial Examples in Environment Perception for Automated Driving (Review)

    Jun Yan and Huilin Yin. Adversarial examples in envi- ronment perception for automated driving. arXiv preprint arXiv:2504.08414, 2025

    work page Pith review arXiv 2025

  5. [5]

    Pointnet: Deep learning on point sets for 3d classification and segmentation

    Charles R Qi, Hao Su, Kaichun Mo, and Leonidas J Guibas. Pointnet: Deep learning on point sets for 3d classification and segmentation. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 652– 660, 2017

    work page 2017

  6. [6]

    V oxelnet: End-to-end learning for point cloud based 3d object detection

    Yin Zhou and Oncel Tuzel. V oxelnet: End-to-end learning for point cloud based 3d object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 4490–4499, 2018

    work page 2018

  7. [7]

    Second: Sparsely embedded convolutional detection

    Yan Yan, Yuxing Mao, and Bo Li. Second: Sparsely embedded convolutional detection. Sensors, 18(10):3337, 2018

    work page 2018

  8. [8]

    Pointpillars: Fast en- coders for object detection from point clouds

    Alex H Lang, Sourabh V ora, Holger Caesar, Lubing Zhou, Jiong Yang, and Oscar Beijbom. Pointpillars: Fast en- coders for object detection from point clouds. In Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12697–12705, 2019

    work page 2019

Show all 49 references
  1. [9]

    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. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 770– 779, 2019

    2019

  2. [10]

    V oxel r-cnn: Towards high performance voxel-based 3d object detection

    Jiajun Deng, Shaoshuai Shi, Peiwei Li, Wengang Zhou, Yanyong Zhang, and Houqiang Li. V oxel r-cnn: Towards high performance voxel-based 3d object detection. In Pro- ceedings of the AAAI conference on artificial intelligence, volume 35, pages 1201–1209, 2021

    2021

  3. [11]

    Center-based 3d object detection and tracking

    Tianwei Yin, Xingyi Zhou, and Philipp Krähenbühl. Center-based 3d object detection and tracking. In 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 11779–11788, 2021. doi: 10.1109/CVPR46437.2021.01161

    2021

  4. [12]

    Graph r-cnn: Towards accurate 3d object detection with semantic-decorated local graph

    Honghui Yang, Zili Liu, Xiaopei Wu, Wenxiao Wang, Wei Qian, Xiaofei He, and Deng Cai. Graph r-cnn: Towards accurate 3d object detection with semantic-decorated local graph. In European conference on computer vision, pages 662–679. Springer, 2022

    2022

  5. [13]

    Enhancing lidar point features with foundation model priors for 3d object detection, 2025

    Yujian Mo, Yan Wu, Junqiao Zhao, Jijun Wang, Yinghao Hu, and Jun Yan. Enhancing lidar point features with foundation model priors for 3d object detection, 2025. URL https://arxiv.org/abs/2507.13899

    2025 arXiv

  6. [14]

    Multi-view 3d object detection network for autonomous driving

    Xiaozhi Chen, Huimin Ma, Ji Wan, Bo Li, and Tian Xia. Multi-view 3d object detection network for autonomous driving. In Proceedings of the IEEE conference on Com- puter Vision and Pattern Recognition, pages 1907–1915, 2017

    1907

  7. [15]

    Frustum pointnets for 3d object detec- tion from rgb-d data

    Charles R Qi, Wei Liu, Chenxia Wu, Hao Su, and Leonidas J Guibas. Frustum pointnets for 3d object detec- tion from rgb-d data. In Proceedings of the IEEE confer- ence on computer vision and pattern recognition , pages 918–927, 2018

    2018

  8. [16]

    Pointpainting: Sequential fusion for 3d object detection

    Sourabh V ora, Alex H Lang, Bassam Helou, and Oscar Beijbom. Pointpainting: Sequential fusion for 3d object detection. In Proceedings of the IEEE /CVF conference on computer vision and pattern recognition, pages 4604– 4612, 2020

    2020

  9. [17]

    Epnet: Enhancing point features with image semantics for 3d object detection

    Tengteng Huang, Zhe Liu, Xiwu Chen, and Xiang Bai. Epnet: Enhancing point features with image semantics for 3d object detection. In Computer vision–ECCV 2020: 16th European conference, Glasgow, UK, August 23–28, 2020, proceedings, part XV 16, pages 35–52. Springer, 2020. Prepr...

    2020

  10. [18]

    Focal sparse convolutional networks for 3d object detection

    Yukang Chen, Yanwei Li, Xiangyu Zhang, Jian Sun, and Jiaya Jia. Focal sparse convolutional networks for 3d object detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5428–5437, 2022

    2022

  11. [19]

    Sparse fuse dense: Towards high quality 3d detection with depth completion

    Xiaopei Wu, Liang Peng, Honghui Yang, Liang Xie, Chenxi Huang, Chengqi Deng, Haifeng Liu, and Deng Cai. Sparse fuse dense: Towards high quality 3d detection with depth completion. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 5418...

    2022

  12. [20]

    Logonet: Towards accurate 3d object detection with local-to-global cross-modal fusion

    Xin Li, Tao Ma, Yuenan Hou, Botian Shi, Yuchen Yang, Youquan Liu, Xingjiao Wu, Qin Chen, Yikang Li, Yu Qiao, et al. Logonet: Towards accurate 3d object detection with local-to-global cross-modal fusion. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern R...

    2023

  13. [21]

    Virtual sparse convolution for multimodal 3d object detection

    Hai Wu, Chenglu Wen, Shaoshuai Shi, Xin Li, and Cheng Wang. Virtual sparse convolution for multimodal 3d object detection. In Proceedings of the IEEE /CVF conference on computer vision and pattern recognition, pages 21653– 21662, 2023

    2023

  14. [22]

    Sparse query dense: Enhancing 3d object detection with pseudo points

    Yujian Mo, Yan Wu, Junqiao Zhao, Zhenjie Hou, Weiquan Huang, Yinghao Hu, Jijun Wang, and Jun Yan. Sparse query dense: Enhancing 3d object detection with pseudo points. In Proceedings of the 32nd ACM International Conference on Multimedia, pages 409–418, 2024

    2024

  15. [23]

    Structure aware single-stage 3d ob- ject detection from point cloud

    Chenhang He, Hui Zeng, Jianqiang Huang, Xian-Sheng Hua, and Lei Zhang. Structure aware single-stage 3d ob- ject detection from point cloud. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11873–11882, 2020

    2020

  16. [24]

    Se-ssd: Self-ensembling single-stage object detector from point cloud

    Wu Zheng, Weiliang Tang, Li Jiang, and Chi-Wing Fu. Se-ssd: Self-ensembling single-stage object detector from point cloud. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14494– 14503, 2021

    2021

  17. [25]

    From points to parts: 3d object detec- tion from point cloud with part-aware and part-aggregation network

    Shaoshuai Shi, Zhe Wang, Jianping Shi, Xiaogang Wang, and Hongsheng Li. From points to parts: 3d object detec- tion from point cloud with part-aware and part-aggregation network. IEEE transactions on pattern analysis and ma- chine intelligence, 43(8):2647–2664, 2020

    2020

  18. [26]

    Joint 3d proposal generation and object detection from view aggregation

    Jason Ku, Melissa Mozifian, Jungwook Lee, Ali Harakeh, and Steven L Waslander. Joint 3d proposal generation and object detection from view aggregation. In 2018 IEEE/RSJ international conference on intelligent robots and systems (IROS), pages 1–8. IEEE, 2018

    2018

  19. [27]

    Pointaugmenting: Cross-modal augmentation for 3d object detection

    Chunwei Wang, Chao Ma, Ming Zhu, and Xiaokang Yang. Pointaugmenting: Cross-modal augmentation for 3d object detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 11794– 11803, 2021

    2021

  20. [28]

    Bevfusion: A simple and robust lidar-camera fusion framework

    Tingting Liang, Hongwei Xie, Kaicheng Yu, Zhongyu Xia, Zhiwei Lin, Yongtao Wang, Tao Tang, Bing Wang, and Zhi Tang. Bevfusion: A simple and robust lidar-camera fusion framework. Advances in Neural Information Processing Systems, 35:10421–10434, 2022

    2022

  21. [29]

    Bevfusion: Multi-task multi-sensor fusion with unified bird’s-eye view representation

    Zhijian Liu, Haotian Tang, Alexander Amini, Xinyu Yang, Huizi Mao, Daniela L Rus, and Song Han. Bevfusion: Multi-task multi-sensor fusion with unified bird’s-eye view representation. In 2023 IEEE international conference on Preprint – LDRFusion: A LiDAR-Dominant multimodal ref...

    2023

  22. [30]

    Mul- timodal virtual point 3d detection

    Tianwei Yin, Xingyi Zhou, and Philipp Krähenbühl. Mul- timodal virtual point 3d detection. Advances in Neural Information Processing Systems, 34:16494–16507, 2021

    2021

  23. [31]

    Msmdfusion: Fusing lidar and camera at multiple scales with multi-depth seeds for 3d object detection

    Yang Jiao, Zequn Jie, Shaoxiang Chen, Jingjing Chen, Lin Ma, and Yu-Gang Jiang. Msmdfusion: Fusing lidar and camera at multiple scales with multi-depth seeds for 3d object detection. In Proceedings of the IEEE /CVF conference on computer vision and pattern recognition , pages ...

    2023

  24. [32]

    Pnp-3d: A plug- and-play for 3d point clouds

    Shi Qiu, Saeed Anwar, and Nick Barnes. Pnp-3d: A plug- and-play for 3d point clouds. IEEE Transactions on Pat- tern Analysis and Machine Intelligence, 45(1):1312–1319, 2021

    2021

  25. [33]

    Dgcnn: A convolutional neural network over large-scale labeled graphs

    Anh Viet Phan, Minh Le Nguyen, Yen Lam Hoang Nguyen, and Lam Thu Bui. Dgcnn: A convolutional neural network over large-scale labeled graphs. Neural Networks, 108: 533–543, 2018

    2018

  26. [34]

    Generalized intersection over union: A metric and a loss for bounding box regression

    Hamid Rezatofighi, Nathan Tsoi, JunYoung Gwak, Amir Sadeghian, Ian Reid, and Silvio Savarese. Generalized intersection over union: A metric and a loss for bounding box regression. In Proceedings of the IEEE/CVF confer- ence on computer vision and pattern recognition , pages 65...

    2019

  27. [35]

    Vision meets robotics: The kitti dataset

    Andreas Geiger, Philip Lenz, Christoph Stiller, and Raquel Urtasun. Vision meets robotics: The kitti dataset. The international journal of robotics research, 32(11):1231– 1237, 2013

    2013

  28. [36]

    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, Giancarlo Baldan, and Oscar Beijbom. nuscenes: A multi- modal dataset for autonomous driving. In Proceedings of the IEEE/CVF conference on computer vision and pattern...

    2020

  29. [37]

    Depth completion with twin surface extrapolation at occlusion boundaries

    Saif Imran, Xiaoming Liu, and Daniel Morris. Depth completion with twin surface extrapolation at occlusion boundaries. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 2583– 2592, 2021

    2021

  30. [38]

    Casa: A cascade attention network for 3-d object detection from lidar point clouds.IEEE Transac- tions on Geoscience and Remote Sensing, 60:1–11, 2022

    Hai Wu, Jinhao Deng, Chenglu Wen, Xin Li, Cheng Wang, and Jonathan Li. Casa: A cascade attention network for 3-d object detection from lidar point clouds.IEEE Transac- tions on Geoscience and Remote Sensing, 60:1–11, 2022

    2022

  31. [39]

    Glenet: Boosting 3d object detectors with generative label uncertainty estimation

    Yifan Zhang, Qijian Zhang, Zhiyu Zhu, Junhui Hou, and Yixuan Yuan. Glenet: Boosting 3d object detectors with generative label uncertainty estimation. International Jour- nal of Computer Vision, 131(12):3332–3352, 2023

    2023

  32. [40]

    3onet: 3-d detector for occluded object under obstructed conditions

    Hiep Anh Hoang and Myungsik Yoo. 3onet: 3-d detector for occluded object under obstructed conditions. IEEE Sensors Journal, 23(16):18879–18892, 2023

    2023

  33. [41]

    Pvt-ssd: Single-stage 3d object detector with point-voxel transformer

    Honghui Yang, Wenxiao Wang, Minghao Chen, Binbin Lin, Tong He, Hua Chen, Xiaofei He, and Wanli Ouyang. Pvt-ssd: Single-stage 3d object detector with point-voxel transformer. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13476– 13487, 2023

    2023

  34. [42]

    Transformation-equivariant 3d object detection for autonomous driving

    Hai Wu, Chenglu Wen, Wei Li, Xin Li, Ruigang Yang, and Cheng Wang. Transformation-equivariant 3d object detection for autonomous driving. In Proceedings of the AAAI Conference on Artificial Intelligence , volume 37, pages 2795–2802, 2023

    2023

  35. [43]

    Sparsedet: a simple and effective framework for fully sparse lidar-based 3d object detection

    Lin Liu, Ziying Song, Qiming Xia, Feiyang Jia, Caiyan Jia, Lei Yang, Yan Gong, and Hongyu Pan. Sparsedet: a simple and effective framework for fully sparse lidar-based 3d object detection. IEEE Transactions on Geoscience and Remote Sensing, 2024

    2024

  36. [44]

    Investigating 3d object detection using stereo camera and lidar fusion with bird’s-eye view representation

    Xin Nie, Lin Zhu, Zhicheng He, Aiguo Cheng, Shengshi Zhong, and Eric Li. Investigating 3d object detection using stereo camera and lidar fusion with bird’s-eye view representation. Neurocomputing, 620:129144, 2025

    2025

  37. [45]

    Sslfusion: Scale and space aligned latent fusion model for multi- modal 3d object detection

    Bonan Ding, Jin Xie, Jing Nie, and Jiale Cao. Sslfusion: Scale and space aligned latent fusion model for multi- modal 3d object detection. In Proceedings of the AAAI Conference on Artificial Intelligence , volume 39, pages 2735–2743, 2025

    2025

  38. [46]

    Vikienet: Towards efficient 3d object detection with virtual key in- stance enhanced network

    Zhuochen Yu, Bijie Qiu, and Andy WH Khong. Vikienet: Towards efficient 3d object detection with virtual key in- stance enhanced network. In Proceedings of the Computer Vision and Pattern Recognition Conference, pages 11844– 11853, 2025

    2025

  39. [47]

    Point density-aware voxels for lidar 3d object detection

    Jordan SK Hu, Tianshu Kuai, and Steven L Waslander. Point density-aware voxels for lidar 3d object detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8469–8478, 2022

    2022

  40. [48]

    V oxel field fusion for 3d object detection

    Yanwei Li, Xiaojuan Qi, Yukang Chen, Liwei Wang, Zem- ing Li, Jian Sun, and Jiaya Jia. V oxel field fusion for 3d object detection. In Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition, pages 1120–1129, 2022

    2022

  41. [49]

    Tsstdet: transformation-based 3-d object detection via a spatial shape transformer

    Hiep Anh Hoang, Duy Cuong Bui, and Myungsik Yoo. Tsstdet: transformation-based 3-d object detection via a spatial shape transformer. IEEE Sensors Journal, 24(5): 7126–7139, 2024

    2024

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.