Pith. sign in

REVIEW 4 major objections 7 minor 54 references

DepthFusion: Depth-Aware Hybrid Feature Fusion for LiDAR-Camera 3D Object Detection

T0 review · 4 major / 7 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Depth-weighted fusion outperforms prior LiDAR-camera 3D detectors

desk verdict Empirically solid, mechanistically underdetermined: the fusion gains look real, but the 'depth-aware' explanation is not actually isolated by the ablations. read the letter →

arxiv 2505.07398 v1 pith:UHV5NGJO submitted 2025-05-12 cs.CV

Mingqian Ji , Jian Yang , Shanshan Zhang This is my paper · ORCID
classification cs.CV
keywords LiDAR-camerafusion3DobjectdetectiondepthencodingBEVfeaturescross-attentionnuScenesKITTIrobustness
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

DepthFusion claims that LiDAR-camera 3D object detection improves when the fusion of point cloud and image features is explicitly conditioned on depth. The paper shows statistically that LiDAR points dominate at close range while images become more informative at long range, and then builds a fusion strategy that lets the model weight each modality by depth. On nuScenes, KITTI, and the corrupted nuScenes-C benchmark, the reported accuracy and robustness exceed previous state-of-the-art methods. If the central claim holds, depth-aware weighting is a simple, parameter-free way to get more out of existing multimodal detectors.

What carries the argument

The depth encoding: a precomputed depth matrix storing the Euclidean distance of each BEV cell from the ego vehicle, converted into a parameter-free sine-cosine embedding following the positional-encoding recipe of the transformer. This embedding is multiplied element-wise into the point cloud BEV features that serve as the query in cross-attention, at both the global scale (Depth-GFusion) and the local instance scale (Depth-LFusion), thereby letting the attention weights modulate image-feature influence by depth. The ablation shows multiplication beats summation or concatenation for embedding the depth signal.

What would settle it

Train the full DepthFusion-base with the depth matrix replaced by a random but fixed permutation of depth values across BEV cells. If NDS stays near the reported 74.0 instead of dropping toward the no-depth ablation level of 71.8, the gains come from the gating machinery, not from depth information specifically.

Watch

Extended reading notes

Core claim

The paper's central claim is that introducing depth encoding into both global BEV fusion and local instance-level fusion lets a LiDAR-camera detector adaptively emphasize whichever modality is more reliable at each distance, improving detection accuracy and robustness. Specifically, Depth-GFusion multiplies a sine-cosine depth encoding into the point cloud BEV query before cross-attending to image BEV features, so image influence grows with depth; Depth-LFusion applies the same depth-adaptive weighting to raw voxel and multi-view image instance features to recover detail lost in BEV projection. The authors report state-of-the-art NDS and mAP on nuScenes (74.9 NDS on validation for the largest variant), gains over prior methods on KITTI, and smaller performance drops under corruptions on nuScenes-C, with the largest accuracy improvements at distances beyond 30 meters.

Load-bearing premise

The extra accuracy from adding depth encoding is caused by the model genuinely modulating modality importance by depth, rather than by the extra cross-attention parameters or the multiplicative gating operation itself.

Editorial extensions

If this is right

  • If correct, depth-adaptive fusion improves detection across all depth ranges, with the largest gains at far range (47.5 mAP beyond 30 meters versus 34.1 for the BEVFusion baseline).
  • The method degrades less under corruptions (a 4.63-point NDS drop on nuScenes-C versus 5.34 for the baseline), suggesting the depth weighting implicitly down-weights unreliable image features.
  • Because the depth encoding is parameter-free, the accuracy gains come with no additional learnable parameters for the depth signal, and the light variant runs at 13.8 FPS while beating the heavier baseline.
  • Combining global and local depth-aware fusion works better than either alone in ablations (73.3 NDS versus 72.4 and 72.7), indicating the two levels carry complementary information.
  • The success on small far-away objects such as traffic cones and barriers suggests depth-aware weighting directly addresses the sparse-point failure mode of LiDAR-only detection.

Reading between the lines

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

  • The same depth-gating principle may transfer to other complementary sensor pairs, such as radar-camera or event-camera fusion, where measurement density also decays with range.
  • A testable extension would be applying only the depth-encoding multiplication to a single-modality detector, isolating whether the gain comes from modality weighting or from the added cross-attention capacity.
  • The robustness result hints that depth gating could be augmented with an explicit per-modality quality estimate, making the weighting respond not just to depth but to weather or sensor degradation.
  • If depth is truly the driving signal, then inverting the depth matrix (making near cells look far) should measurably degrade near-range detection; this prediction is a direct check of the paper's explanatory claim.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 7 minor

Summary. The paper proposes DepthFusion, a LiDAR-camera 3D object detection method built on the BEVFusion baseline. It introduces a parameter-free sine/cosine depth encoding of the BEV coordinates and uses it in two fusion modules: Depth-GFusion (DGF), which multiplies the query in a global cross-attention module by the depth encoding before querying image BEV features, and Depth-LFusion (DLF), which applies a similar depth-scaled attention to locally cropped raw voxel and image features. Experiments on nuScenes, KITTI, and nuScenes-C report consistent improvements over previous multimodal detectors when comparisons are grouped by 2D backbone, and ablations show that each module and the depth encoding contribute to the gains. The central claim is that depth-aware modulation of modality weights, rather than the added fusion capacity alone, is responsible for the improvements, especially for distant objects and under corruptions.

Significance. If the mechanism claim were fully supported, this would be a useful and practical contribution: a parameter-free depth signal that improves a strong baseline across multiple backbones, with clear benefits for far-range detection and robustness to corruptions. The paper's strengths include backbone-matched comparisons on nuScenes and KITTI, ablation of each module, evaluation on the nuScenes-C robustness benchmark, and a parameter-free depth encoding with no extra parameters. The main weakness is that the evidence for the depth-aware mechanism is partly underdetermined: the key ablation removes the entire multiplicative depth interaction, and the robustness explanation is based on post-hoc attention statistics rather than on a property enforced by the architecture. The absence of error bars and code release also limits confidence, though these are common in this line of work.

major comments (4)
  1. [§III-B2, Eq. (3), Table IX] The w/o D ablation does not isolate depth-dependent modulation. In Eq. (3), the query is (V_B^G + P) · D, so removing D simultaneously removes (i) depth-dependent reweighting, (ii) the multiplicative feature transformation itself, and (iii) any fixed gating effect that D may induce on feature scales. The comparison in Table IX therefore conflates at least three mechanisms. Table X shows that multiplication beats summation and concatenation, but that result does not separate these mechanisms either, because multiplication may win simply through a nonlinear scaling effect that does not track depth. To support the paper's central explanatory claim, please add a control experiment in which D is replaced by a depth-independent gate (e.g., a scalar learned from the query features without depth, or a shuffled/random depth map) so that the accuracy gain can be attributed specifically to depth-consistent modulation.
  2. [§IV-D, §IV-J, Fig. 8] The robustness explanation is post-hoc. The model has no corruption-awareness input, so the assertion in Sec. IV-J that 'the degradation of image feature quality leads to a reduction in the weight assigned to the image modality' is not forced by any architectural property; Fig. 8(a) shows attention statistics under different weather conditions but does not establish a causal link. Since the abstract and Sec. IV-D claim superior robustness as a contribution, this mechanism evidence is load-bearing. A stronger test would be to corrupt only the image branch or only the LiDAR branch during inference and measure whether the depth-dependent attention weights shift in the predicted direction, or to compare against a variant in which the depth gating is frozen when corrupted inputs are encountered.
  3. [§IV-E, Table V] The far-distance comparison is internally inconsistent. The text says 'compared to BEVFusion [20]' and 'surpassing ObjectFusion by 13.4 pp', but Table V does not list ObjectFusion and the baseline shown is BEVFusion [21]. Please correct the citation and either add ObjectFusion's per-depth numbers or remove the unsupported margin. In addition, the protocol for depth-stratified mAP (how predictions are assigned to depth groups, and how matching is performed) should be specified so that the far-distance claim in Tables IV and V is reproducible.
  4. [General] No error bars or multi-seed results are reported, and no code is released. For a paper whose main quantitative claim is a few percentage points of NDS/mAP, a single run per configuration makes it difficult to judge whether the reported gaps are outside run-to-run noise. Reporting mean and standard deviation over at least three seeds for the key comparisons and ablations would substantially strengthen the empirical claim.
minor comments (7)
  1. [Fig. 1 and Sec. I] The text reports 163.7 points per object within 0-10 meters, while the figure displays 163.9; please unify the numbers.
  2. [§IV-G] The sentence 'our DepthFusion-light our DepthFusion-light achieves 33.6% mAP' contains a duplicated phrase.
  3. [§III-C] The phrase 'local feauture fusion' contains a typo; it should be 'local feature fusion'.
  4. [Table I] The CMT row appears to have a formatting issue: the columns after FPS contain only three values, whereas other rows contain four; please verify and align the entries.
  5. [Fig. 8] The axes and units of the attention-weight curves are not defined. Please specify what is averaged, over which dataset split, and how 'depth' is discretized.
  6. [§IV-H and Table VII] The row labeled 'BEVFusion + Ours' is unclear and does not correspond to a named model variant; please clarify which configuration this represents.
  7. [Abstract and Sec. I] The claim of being 'the first to observe' that modality roles vary with depth is stated without qualification; please discuss existing depth-aware fusion works (e.g., MSMDFusion's multi-depth seeds, GAFusion's guidance, 3D-DFM's dynamic filters) to clarify the precise novelty.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: DepthFusion's depth-aware fusion is evaluated on external benchmarks with parameter-free depth encoding and module ablations.

full rationale

The paper's central claim is an empirical architecture result: inserting depth-encoded multiplicative gating into global and local cross-attention fusion improves 3D detection on nuScenes, KITTI, and nuScenes-C. The depth encoding D is not fitted to any detection output or benchmark metric; it is a precomputed sine/cosine transformation of Euclidean distances in a fixed BEV grid (Eqs. 1-2), which is parameter-free by the paper's own report. The predicted quantities (NDS, mAP, per-depth mAP, corruption robustness) come from held-out public benchmarks, not from any fitted parameter that was tuned to those benchmarks. Ablations in Tables VIII-X isolate the contribution of each module, the depth encoding, and the multiplication operation, and the comparisons are against external baselines such as BEVFusion. The statistical observation that point clouds become sparse and images relatively informative at long range motivates the design but is not used as a training target or as a constraint on the learned attention weights, so the later attention-weight visualization is explanatory rather than definitional. The skeptical concern that the w/o-D ablation removes the whole multiplicative depth pathway and therefore cannot fully isolate 'depth-aware modulation' from extra nonlinear gating capacity is an attribution or interpretability limitation, not a circular reduction of a prediction to its inputs: no equation in the paper equates the reported gains to the depth encoding by construction, and no load-bearing step reduces to a self-citation. The authors do not invoke any prior uniqueness theorem from themselves, and the few citations to prior fusion work are standard external baselines rather than self-supporting premises. Accordingly, the paper is self-contained against external benchmarks and warrants a circularity score of 0.

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

The paper introduces no new physical entities and no hand-fitted constants; the depth encoding is a fixed sine/cosine function. Its central claim rests on learned network weights and empirical evaluation, so the main assumptions are domain assumptions about dataset representativeness and the interpretation of attention.

assumptions (3)
  • domain assumption The nuScenes-mini statistical analysis with 10 scenes is representative of the full nuScenes and KITTI distributions.
    Section I uses nuScenes-mini to draw conclusions about depth-dependent modality importance that motivate the method and are used to explain results on full datasets.
  • domain assumption The Euclidean distance from ego center in BEV space is an appropriate depth proxy for modulating fusion weights.
    Eq. (2) defines depth as radial BEV distance; the method assumes this correlates with actual object distance and with the information content of each modality.
  • domain assumption The attention weights in the trained DGF/DLF models reflect the intended modality-importance modulation.
    Section IV-J interprets attention maps as evidence that image weight increases with depth, but no training constraint enforces this.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DepthFusion: Depth-Aware Hybrid Feature Fusion for LiDAR-Camera 3D Object Detection." pith.science (2026). https://pith.science/paper/UHV5NGJO

@misc{pith2026250507398,
  author       = {Pith},
  title        = {Pith review of: DepthFusion: Depth-Aware Hybrid Feature Fusion for LiDAR-Camera 3D Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UHV5NGJO}},
  note         = {Machine review of arXiv:2505.07398}
}
read the original abstract

State-of-the-art LiDAR-camera 3D object detectors usually focus on feature fusion. However, they neglect the factor of depth while designing the fusion strategy. In this work, we are the first to observe that different modalities play different roles as depth varies via statistical analysis and visualization. Based on this finding, we propose a Depth-Aware Hybrid Feature Fusion (DepthFusion) strategy that guides the weights of point cloud and RGB image modalities by introducing depth encoding at both global and local levels. Specifically, the Depth-GFusion module adaptively adjusts the weights of image Bird's-Eye-View (BEV) features in multi-modal global features via depth encoding. Furthermore, to compensate for the information lost when transferring raw features to the BEV space, we propose a Depth-LFusion module, which adaptively adjusts the weights of original voxel features and multi-view image features in multi-modal local features via depth encoding. Extensive experiments on the nuScenes and KITTI datasets demonstrate that our DepthFusion method surpasses previous state-of-the-art methods. Moreover, our DepthFusion is more robust to various kinds of corruptions, outperforming previous methods on the nuScenes-C dataset.

Figures

Figures reproduced from arXiv: 2505.07398 by the authors.

Figure 1
Figure 1. Statistical analysis on the nuScenes-mini dataset. The [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Visualization analysis on the nuScenes-mini dataset. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Overview of our method. It introduces depth encoding in both global and local feature fusion to obtain depth-adaptive [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Illustration of the DGF. capture the depth-based weight variations when querying the image BEV features, reflecting how the importance of image features changes with depth. The image BEV features are then queried as the corresponding key and value. We utilize the multi…
Figure 5
Figure 5. Figure 5: Illustration of the DLF. we utilize an RPN to regress t 3D boxes based on the BEV features F G B . We directly crop the global fused BEV features F G B based on the regressed 3D boxes to obtain the local fused BEV features F L B ∈ R c×t . On the other hand, we project …
Figure 6
Figure 6. Figure 6: Qualitative detection results and BEV features of BEVFusion and ours. We show the ground truth boxes in [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Trade-off between performance (NDS) and computa [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Statistics of attention weights and the visualization of [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

54 extracted references · 48 canonical work pages

  1. [20]

    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

    work page 2022

  2. [21]

    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 Robotics and Automation, pages 2774–2781. IEEE, 2023

    work page 2023

  3. [1]

    nuscenes: A multimodal 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 multimodal dataset for autonomous driving. In Conference on Computer Vision and Pattern Recognition, pages 11621– 11631, 2020

    work page 2020

  4. [2]

    Are we ready for autonomous driving? the kitti vision benchmark suite

    Andreas Geiger, Philip Lenz, and Raquel Urtasun. Are we ready for autonomous driving? the kitti vision benchmark suite. In Conference on Computer Vision and Pattern Recognition , pages 3354–3361, 2012

    work page 2012

  5. [3]

    Benchmarking robustness of 3d object detection to common corruptions

    Yinpeng Dong, Caixin Kang, Jinlai Zhang, Zijian Zhu, Yikai Wang, Xiao Yang, Hang Su, Xingxing Wei, and Jun Zhu. Benchmarking robustness of 3d object detection to common corruptions. In Conference on Computer Vision and Pattern Recognition , pages 1022–1032, 2023

    work page 2023

  6. [4]

    Pointrcnn: 3d object proposal generation and detection from point cloud

    Shaoshuai Shi, Xiaogang Wang, and Hongsheng Li. Pointrcnn: 3d object proposal generation and detection from point cloud. In Conference on Computer Vision and Pattern Recognition , pages 770–779, 2019

    work page 2019

  7. [5]

    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

  8. [6]

    Sp-det: Leveraging saliency prediction for voxel-based 3d object detection in sparse point cloud

    Pei An, Yucong Duan, Yuliang Huang, Jie Ma, Yanfei Chen, Liheng Wang, You Yang, and Qiong Liu. Sp-det: Leveraging saliency prediction for voxel-based 3d object detection in sparse point cloud. IEEE Transactions on Multimedia , 2023

    work page 2023

Show all 54 references
  1. [7]

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

    Shaoshuai Shi, Chaoxu Guo, Li Jiang, Zhe Wang, Jianping Shi, Xi- aogang Wang, and Hongsheng Li. Pv-rcnn: Point-voxel feature set abstraction for 3d object detection. In Conference on Computer Vision and Pattern Recognition, pages 10529–10538, 2020

    2020

  2. [8]

    Fcos3d: Fully convolutional one-stage monocular 3d object detection

    Tai Wang, Xinge Zhu, Jiangmiao Pang, and Dahua Lin. Fcos3d: Fully convolutional one-stage monocular 3d object detection. In International Conference on Computer Vision , pages 913–922, 2021

    2021

  3. [9]

    Bevdet: High-performance multi-camera 3d object detection in bird-eye- view

    Junjie Huang, Guan Huang, Zheng Zhu, Yun Ye, and Dalong Du. Bevdet: High-performance multi-camera 3d object detection in bird-eye- view. arXiv preprint arXiv:2112.11790 , 2021

    2021 arXiv

  4. [10]

    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 Conference on Computer Vision and Pattern Recognition , pages 4604–4612, 2020

    2020

  5. [11]

    Pointaug- menting: Cross-modal augmentation for 3d object detection

    Chunwei Wang, Chao Ma, Ming Zhu, and Xiaokang Yang. Pointaug- menting: Cross-modal augmentation for 3d object detection. In Confer- ence on Computer Vision and Pattern Recognition, pages 11794–11803, 2021

    2021

  6. [12]

    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 Conference on Computer Vision and Pattern Recognition , pages 21653–21662, 2023

    2023

  7. [13]

    Virpnet: A multimodal virtual point generation network for 3d object detection

    Lin Wang, Shiliang Sun, and Jing Zhao. Virpnet: A multimodal virtual point generation network for 3d object detection. IEEE Transactions on Multimedia, 2024

    2024

  8. [14]

    Transfusion: Robust lidar-camera fusion for 3d object detection with transformers

    Xuyang Bai, Zeyu Hu, Xinge Zhu, Qingqiu Huang, Yilun Chen, Hongbo Fu, and Chiew-Lan Tai. Transfusion: Robust lidar-camera fusion for 3d object detection with transformers. In Conference on Computer Vision and Pattern Recognition, pages 1090–1099, 2022

    2022

  9. [15]

    Vpfnet: Improving 3d object detection with virtual point based lidar and stereo data fusion

    Hanqi Zhu, Jiajun Deng, Yu Zhang, Jianmin Ji, Qiuyu Mao, Houqiang Li, and Yanyong Zhang. Vpfnet: Improving 3d object detection with virtual point based lidar and stereo data fusion. IEEE Transactions on Multimedia, 2022

    2022

  10. [16]

    Futr3d: A unified sensor fusion framework for 3d detection

    Xuanyao Chen, Tianyuan Zhang, Yue Wang, Yilun Wang, and Hang Zhao. Futr3d: A unified sensor fusion framework for 3d detection. In Conference on Computer Vision and Pattern Recognition , pages 172– 181, 2023

    2023

  11. [17]

    Deformable detr: Deformable transformers for end-to-end object detection

    Xizhou Zhu, Weijie Su, Lewei Lu, Bin Li, Xiaogang Wang, and Jifeng Dai. Deformable detr: Deformable transformers for end-to-end object detection. arXiv preprint arXiv:2010.04159 , 2020

    2010 arXiv

  12. [18]

    3d-dfm: Anchor-free multimodal 3-d object detection with dynamic fusion module for autonomous driving

    Chunmian Lin, Daxin Tian, Xuting Duan, Jianshan Zhou, Dezong Zhao, and Dongpu Cao. 3d-dfm: Anchor-free multimodal 3-d object detection with dynamic fusion module for autonomous driving. IEEE Transactions on Neural Networks and Learning Systems , 34(12):10812–10822, 2022

    2022

  13. [19]

    Cl3d: Camera-lidar 3d object detection with point feature enhancement and point-guided fusion

    Chunmian Lin, Daxin Tian, Xuting Duan, Jianshan Zhou, Dezong Zhao, and Dongpu Cao. Cl3d: Camera-lidar 3d object detection with point feature enhancement and point-guided fusion. IEEE Transactions on Intelligent Transportation Systems, 23(10):18040–18050, 2022

    2022

  14. [22]

    Sparsefusion: Fusing multi-modal sparse representations for multi- sensor 3d object detection

    Yichen Xie, Chenfeng Xu, Marie-Julie Rakotosaona, Patrick Rim, Federico Tombari, Kurt Keutzer, Masayoshi Tomizuka, and Wei Zhan. Sparsefusion: Fusing multi-modal sparse representations for multi- sensor 3d object detection. In International Conference on Computer Vision, pages...

    2023

  15. [23]

    Ob- jectfusion: Multi-modal 3d object detection with object-centric fusion

    Qi Cai, Yingwei Pan, Ting Yao, Chong-Wah Ngo, and Tao Mei. Ob- jectfusion: Multi-modal 3d object detection with object-centric fusion. In International Conference on Computer Vision , pages 18067–18076, 2023

    2023

  16. [24]

    Cross modal transformer via coordinates encoding for 3d object dectection

    Junjie Yan, Yingfei Liu, Jianjian Sun, Fan Jia, Shuailin Li, Tiancai Wang, and Xiangyu Zhang. Cross modal transformer via coordinates encoding for 3d object dectection. arXiv preprint arXiv:2301.01283 , 2023

    2023 arXiv

  17. [25]

    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 Conference on Computer Vision and Pattern Recognition, pages 17524– 17534, 2023

    2023

  18. [26]

    Is-fusion: Instance-scene collaborative fusion for multimodal 3d object 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 object detection. In Conference on Computer Vision and Pattern Recognition , pages 14905–14915, 2024

    2024

  19. [27]

    Clocs: Camera-lidar object candidates fusion for 3d object detection

    Su Pang, Daniel Morris, and Hayder Radha. Clocs: Camera-lidar object candidates fusion for 3d object detection. In 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems , pages 10386–10393. IEEE, 2020

    2020

  20. [28]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems , 30, 2017

    2017

  21. [29]

    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 Proceedings of the AAAI Conference on Artificial Intelligence, pages 1201–1209, 2021

    2021

  22. [30]

    Mask r-cnn

    Kaiming He, Georgia Gkioxari, Piotr Dollar, and Ross Girshick. Mask r-cnn. IEEE transactions on pattern analysis and machine intelligence , 42(2):386–397, 2020

    2020

  23. [31]

    MMDetection3D: OpenMMLab next- generation platform for general 3D object detection

    MMDetection3D Contributors. MMDetection3D: OpenMMLab next- generation platform for general 3D object detection. https://github.com/ open-mmlab/mmdetection3d, 2020

    2020

  24. [32]

    Second: Sparsely embedded convolutional detection

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

    2018

  25. [33]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 770– 778, 2016

    2016

  26. [34]

    A convnet for the 2020s

    Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell, and Saining Xie. A convnet for the 2020s. In Conference on Computer Vision and Pattern Recognition, pages 11976–11986, 2022

    2022

  27. [35]

    Feature pyramid networks for object detection

    Tsung-Yi Lin, Piotr Doll ´ar, Ross Girshick, Kaiming He, Bharath Hariha- ran, and Serge Belongie. Feature pyramid networks for object detection. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 2117–2125, 2017

    2017

  28. [36]

    Bevpoolv2: A cutting-edge implemen- tation of bevdet toward deployment

    Junjie Huang and Guan Huang. Bevpoolv2: A cutting-edge implemen- tation of bevdet toward deployment. arXiv:2211.17111, 2022

    2022 arXiv

  29. [37]

    Deepinteraction: 3d object detection via modality interaction

    Zeyu Yang, Jiaqi Chen, Zhenwei Miao, Wei Li, Xiatian Zhu, and Li Zhang. Deepinteraction: 3d object detection via modality interaction. Advances in Neural Information Processing Systems , 35:1992–2005, 2022

    1992

  30. [38]

    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 Conference on Computer Vision and Pattern Recognition , pages 21643–21652, 2023

    2023

  31. [39]

    Focalformer3d: focusing on hard instance for 3d object detection

    Yilun Chen, Zhiding Yu, Yukang Chen, Shiyi Lan, Anima Anandkumar, Jiaya Jia, and Jose M Alvarez. Focalformer3d: focusing on hard instance for 3d object detection. In International Conference on Computer Vision, pages 8394–8405, 2023

    2023

  32. [40]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. In International Conference on Computer Vision, pages 10012–10022, 2021

    2021

  33. [41]

    Gafusion: Adaptive fusing lidar and camera with multiple guidance for 3d object detection

    Xiaotian Li, Baojie Fan, Jiandong Tian, and Huijie Fan. Gafusion: Adaptive fusing lidar and camera with multiple guidance for 3d object detection. In Conference on Computer Vision and Pattern Recognition , pages 21209–21218, 2024

    2024

  34. [42]

    Deformable feature aggregation for dynamic multi- modal 3d object detection

    Zehui Chen, Zhenyu Li, Shiquan Zhang, Liangji Fang, Qinhong Jiang, and Feng Zhao. Deformable feature aggregation for dynamic multi- modal 3d object detection. In Proceedings of the European Conference on Computer Vision , pages 628–644. Springer, 2022

    2022

  35. [43]

    Cspnet: A new backbone that can enhance learning capability of cnn

    Chien-Yao Wang, Hong-Yuan Mark Liao, Yueh-Hua Wu, Ping-Yang Chen, Jun-Wei Hsieh, and I-Hau Yeh. Cspnet: A new backbone that can enhance learning capability of cnn. In Conference on Computer Vision and Pattern Recognition Workshops , pages 390–391, 2020

    2020

  36. [44]

    Unifying voxel-based representation with transformer for 3d object detection

    Yanwei Li, Yilun Chen, Xiaojuan Qi, Zeming Li, Jian Sun, and Jiaya Jia. Unifying voxel-based representation with transformer for 3d object detection. Advances in Neural Information Processing Systems , 35:18442–18455, 2022

    2022

  37. [45]

    An energy and gpu-computation efficient backbone network for real-time object detection

    Youngwan Lee, Joong-won Hwang, Sangrok Lee, Yuseok Bae, and Jongyoul Park. An energy and gpu-computation efficient backbone network for real-time object detection. In Conference on Computer Vision and Pattern Recognition Workshops , pages 0–0, 2019

    2019

  38. [46]

    Unitr: A unified and efficient multi- modal transformer for bird’s-eye-view representation

    Haiyang Wang, Hao Tang, Shaoshuai Shi, Aoxue Li, Zhenguo Li, Bernt Schiele, and Liwei Wang. Unitr: A unified and efficient multi- modal transformer for bird’s-eye-view representation. In International Conference on Computer Vision , pages 6792–6802, 2023

    2023

  39. [47]

    Dsvt: Dynamic sparse voxel transformer with rotated sets

    Haiyang Wang, Chen Shi, Shaoshuai Shi, Meng Lei, Sen Wang, Di He, Bernt Schiele, and Liwei Wang. Dsvt: Dynamic sparse voxel transformer with rotated sets. In Conference on Computer Vision and Pattern Recognition, pages 13520–13529, 2023

    2023

  40. [48]

    Unipad: A universal pre-training paradigm for autonomous driving

    Honghui Yang, Sha Zhang, Di Huang, Xiaoyang Wu, Haoyi Zhu, Tong He, Shixiang Tang, Hengshuang Zhao, Qibo Qiu, Binbin Lin, et al. Unipad: A universal pre-training paradigm for autonomous driving. In Conference on Computer Vision and Pattern Recognition, pages 15238– 15250, 2024

    2024

  41. [49]

    Sparselif: High-performance sparse lidar-camera fusion for 3d object detection

    Hongcheng Zhang, Liu Liang, Pengxin Zeng, Xiao Song, and Zhe Wang. Sparselif: High-performance sparse lidar-camera fusion for 3d object detection. In Proceedings of the European Conference on Computer Vision, pages 109–128. Springer, 2024

    2024

  42. [50]

    Pointfusion: Deep sensor fusion for 3d bounding box estimation

    Danfei Xu, Dragomir Anguelov, and Ashesh Jain. Pointfusion: Deep sensor fusion for 3d bounding box estimation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 244–253, 2018

    2018

  43. [51]

    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 Proceedings of the European Conference on Computer Vision , pages 35–52. Springer, 2020

    2020

  44. [52]

    Cat-det: Contrastively aug- mented transformer for multi-modal 3d object detection

    Yanan Zhang, Jiaxin Chen, and Di Huang. Cat-det: Contrastively aug- mented transformer for multi-modal 3d object detection. In Conference on Computer Vision and Pattern Recognition , pages 908–917, 2022

    2022

  45. [53]

    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 Conference on Computer Vision and Pattern Recognition , pages 5428–5437, 2022

    2022

  46. [54]

    V oxel field fusion for 3d object detection

    Yanwei Li, Xiaojuan Qi, Yukang Chen, Liwei Wang, Zeming Li, Jian Sun, and Jiaya Jia. V oxel field fusion for 3d object detection. In Conference on Computer Vision and Pattern Recognition , pages 1120– 1129, 2022

    2022

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.