Pith. sign in

REVIEW 4 major objections 5 minor 34 references

Reliability-Driven LiDAR-Camera Fusion for Robust 3D Object Detection

T0 review · 4 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read ReliFusion fuses LiDAR and camera with per-sensor confidence weights so 3D detection survives sensor failures.

desk verdict Useful robustness study undermined by an inconsistent baseline protocol and an unvalidated confidence module; the core idea is worth a serious look but the current evidence does not support the SOTA robustness claim. read the letter →

arxiv 2502.01856 v1 pith:WF674QDP submitted 2025-02-03 cs.CV cs.LG

Reza Sadeghian , Niloofar Hooshyaripour , Chris Joslin , WonSook Lee This is my paper · ORCID
classification cs.CVcs.LG
keywords 3DobjectdetectionLiDAR-camerafusionsensorreliabilitybird's-eyeviewcross-modalitycontrastivelearningconfidence-weightedcross-attentiontemporalfeatureaggregationnuScenes
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

ReliFusion is a LiDAR-camera fusion framework for 3D object detection in autonomous driving. It is built around the idea that a fusion model should know when one of its sensors is unreliable: the network estimates a confidence score for the LiDAR stream and for the camera stream, then scales each modality's contribution to a mutual cross-attention step by that score. The authors train these confidence scores with contrastive learning that pulls clean LiDAR-camera pairs together and pushes corrupted pairs apart. On the nuScenes benchmark they report accuracy competitive with or above current fusion methods on clean data, and clearly higher mAP than BEVFusion when the LiDAR field of view is narrowed or removed entirely, as well as under camera failure and occlusion. The practical target is graceful degradation: a blinded or disconnected sensor should reduce detection quality smoothly instead of collapsing the detector.

What carries the argument

The central mechanism is the confidence-weighted mutual cross-attention (CW-MCA) fusion rule, fed by the contrastively trained Reliability module. In CW-MCA the camera features query the LiDAR features and the LiDAR features query the camera features; each attention output is multiplied by a scalar confidence score, $C_{\text{Camera}}$ or $C_{\text{LiDAR}}$, produced by a sigmoid over a learned projection of that modality's contrastive embedding. The two confidence-scaled streams are added to form the fused BEV feature that goes into the detection head. Before fusion, the Spatio-Temporal Feature Aggregation (STFA) module applies self-attention across the six camera views and across time steps, so the image stream has a stable multi-frame representation that can carry detection when LiDAR is reduced.

What would settle it

Re-run the Table 2 and Table 3 comparisons with BEVFusion and TransFusion using ReliFusion's backbone, training schedule, temporal frames, and corruption pipeline; if either baseline matches or exceeds ReliFusion's mAP under limited LiDAR FOV, zero LiDAR, or camera occlusion, the reliability-driven advantage is not confirmed.

Watch

Extended reading notes

Core claim

The paper's central claim is that the brittleness of existing LiDAR-camera fusion under sensor malfunction comes from fixed fusion weighting, and that replacing it with confidence-weighted fusion restores robustness. ReliFusion's Reliability module embeds LiDAR and camera BEV features in a shared space, uses cross-modality contrastive learning to make clean pairs agree and corrupted pairs disagree, and maps each embedding to a sigmoid confidence score. The Confidence-Weighted Mutual Cross-Attention module then computes attention in both directions between the modalities and multiplies each direction's output by the corresponding confidence, so degraded input is down-weighted before features are summed. With a LiDAR field of view truncated to nothing, the authors report 24.6 mAP against 12.4 for BEVFusion; with the front camera missing or object pixels occluded, the drop from clean performance is a few mAP points. The conclusion they draw is that the same framework can serve both clean and degraded sensing without a separate failure-detection network.

Load-bearing premise

The robustness comparison assumes that the baseline methods were run under exactly the same protocol as ReliFusion, meaning the same backbones, the same number of temporal frames, and the same corruption simulation.

Editorial extensions

If this is right

  • Narrowing the LiDAR field of view degrades ReliFusion's detection gradually, with reported mAP of 52.4 at $(-\pi/2, \pi/2)$, 44.9 at $(-\pi/3, \pi/3)$, and 24.6 when no LiDAR remains.
  • Dropping the front camera or occluding half the object pixels costs only a few mAP points, since the LiDAR geometry still anchors the fused representation.
  • The reliability scores are trained end-to-end with detection, contrastive, temporal, and confidence losses, so no separate failure classifier is needed at runtime.
  • Under the authors' ablations, each added component -- STFA, CW-MCA, and reliability scoring -- raises limited-FOV and object-failure performance, with the largest relative gains appearing when LiDAR is most degraded.

Reading between the lines

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

  • A testable extension the paper does not run: attach the reliability module and CW-MCA to other BEV fusion backbones; if their degraded-LiDAR mAP rises by the same margin, the weighting mechanism rather than the specific backbone is the active ingredient.
  • The confidence scores could be exported to downstream driving modules as a per-frame trust signal, letting a planner slow down or switch strategies when the camera becomes the only reliable modality.
  • Because the contrastive head was trained on clean-versus-corrupted pairs from this task's corruption types, it is an open question whether unseen corruptions such as rain, snow, or lens dirt also separate from the clean cluster; that is directly measurable with the paper's pipeline.
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 / 5 minor

Summary. The manuscript proposes ReliFusion, a BEV-space LiDAR-camera fusion detector for nuScenes. Three components are introduced: STFA (spatial and temporal attention over camera BEV features), a Reliability module built on cross-modality contrastive learning that outputs per-modality confidence scores, and CW-MCA that weights cross-attention fusion by these scores. Experiments on the nuScenes test set report 70.6 mAP / 73.2 NDS, and robustness experiments under limited LiDAR FOV and object-region point drops report higher mAP than BEVFusion and other baselines. Ablation tables decompose the contributions of STFA, CW-MCA, and the Reliability module. The authors conclude that ReliFusion adaptively balances modalities and is robust to sensor malfunctions.

Significance. If the empirical claims were properly controlled, ReliFusion would address a practically important failure mode: maintaining 3D detection performance under partial or complete LiDAR loss. The architecture is plausible and the ablation design is useful. However, the paper currently does not establish its central claim. The robustness comparison in Tables 2 and 3 is invalidated by unspecified and inconsistent baseline protocols; the supervision target for the confidence loss is undefined; and the contrastive training data for the reliability module is not described. These are load-bearing gaps rather than presentation issues. The paper would benefit from releasing code and checkpoints to enable verification.

major comments (4)
  1. [Section 4.4, Tables 2 and 3] The baseline protocol is under-specified and the clean numbers are inconsistent with Table 1. For example, CenterPoint drops from 60.3 to 56.8 mAP, TransFusion from 68.9 to 66.9, BEVFusion from 69.2 to 67.9, and PointAugmenting from 66.8 to 46.9 between Table 1 and Tables 2/3. Section 4.2 details only ReliFusion's implementation (ConvMixer, VoxelNet, 448x800 images), and no statement is made about which backbones, temporal windows, or corruption simulation were used for the baselines. Since the STFA module (Section 3.2) aggregates camera features over T frames while the baselines are not described as temporal, the limited-FOV margins (52.4 vs 46.4 at (-pi/2,pi/2), 24.6 vs 12.4 with zero LiDAR) could be due to temporal camera information or training settings rather than to the reliability-driven fusion. The central robustness claim is therefore not established.
  2. [Sections 3.3 and 3.5, Eq. (16)] The confidence loss L_conf has no defined supervision target. The scores C_LiDAR and C_Camera are defined in Eq. (12) as sigmoid outputs of learned projections, and L_conf is described only as 'penalizes inaccurate confidence predictions,' but no ground-truth confidence label is defined anywhere in the paper. The first-stage 'regression loss' for pre-training the Confidence Module is likewise unspecified. Without a target, the reliability scores may simply be hidden parameters that improve the detection objective, and the claim that they 'quantify the dependability of each modality' is circular rather than validated. The authors should define the target (for example, the degree of injected corruption) and show that the scores track corruption severity.
  3. [Section 3.3, Eq. (11)] The contrastive learning setup is incompletely specified. Positive pairs are said to be 'unaltered data' and negative pairs 'corrupted data,' but Section 4.2 does not describe any corruption augmentation used during training, nor does it state whether the robustness-test corruptions are seen in pretraining. If the model is trained only on clean data, the negatives in Eq. (11) are other camera embeddings in the batch and the loss does not learn a reliable/corrupted distinction. A concrete training-time corruption protocol is required for the reliability module to be meaningful.
  4. [Tables 2 and 4] The same experimental condition is reported with different numbers. For zero LiDAR FOV, Table 2 gives ReliFusion 24.6/28.7 (mAP/NDS), while Table 4 gives 24.6/39.7; for (-pi/3,pi/3), Table 2 gives 44.9/54.6 and Table 4 gives 44.9/54.8. The NDS differences are unexplained and make the ablation comparisons internally inconsistent. Please reconcile the reported numbers.
minor comments (5)
  1. [Section 4.4, Table 3] The camera-failure protocol is ambiguous: the text says the front (F) image was 'entirely removed or preserved,' but it does not state what happens to the other five cameras in each condition; the 'Preserve F' condition should be defined explicitly.
  2. [Table 3] There is a typographical error: '47.8.0/59.4' for MVX-Net under Object Failure should presumably be '47.8/59.4'.
  3. [Section 4.4, Tables 2 and 4] The notation '(-0, 0)' is confusing; it should be written as the empty set or as '(0,0)' with a clear definition of the zero-FOV protocol.
  4. [Table 2] The heading 'Object Detection Failures' is misleading; the experiment drops points inside object bounding boxes, so a heading such as 'LiDAR Object-Region Drop' would be more accurate.
  5. [Section 4.2] The paper credits MMDetection3D to reference [32], which is the MMDetection paper; either use the correct framework reference or clarify the relationship.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found: the reliability scores are learned intermediate weights, and the central claims rest on external nuScenes comparisons rather than on the paper's own definitions.

full rationale

ReliFusion is an empirical systems paper whose central claim is an evaluated performance difference on nuScenes, not a first-principles derivation. The confidence scores in Eq. (12) are intermediate learned weights produced by an MLP from embeddings; they are not defined in terms of the detection output, and no equation in the paper reduces a reported prediction to a fitted input. The concern that the confidence module is never compared with ground-truth reliability labels is a measurement and interpretation limitation, not a circular step: the scores could in principle be miscalibrated while still improving end-task accuracy, and the ablation in Table 4 gives independent evidence that the module contributes to the reported mAP. The inconsistency between clean baseline numbers in Table 1 versus Tables 2 and 3 is a protocol-validity concern about how baselines were re-run, not a circularity of the proposed method. The only self-citation is reference [33], used to justify the ConvMixer 2D backbone in Section 4.2; it is a design choice and is not load-bearing for the reliability or robustness claims. Therefore the paper does not exhibit a circular derivation chain, and the appropriate circularity score is 0.

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

The central claim rests on hand-chosen hyperparameters and on assumptions about corruptibility separability and the detection head's operation without LiDAR. No external validation of the confidence scores is provided.

free parameters (3)
  • loss weights lambda_1 to lambda_4 = 1.0, 0.1, 0.2, 0.05
    Chosen by hand in Eq. 16; no sensitivity analysis is reported.
  • contrastive temperature tau = 0.07
    Set in implementation details without justification or ablation.
  • embedding size = 128
    Chosen for the contrastive projectors; no ablation is reported.
assumptions (3)
  • domain assumption Corrupted and clean modality features are separable in the shared embedding space
    The CMCL loss (Eq. 11) relies on negative pairs being corrupted data, but the paper does not define how corrupted pairs are generated during training or whether real corruptions follow the same distribution.
  • domain assumption The TransFusion detection head can generate detections from camera-only BEV features when LiDAR input is empty
    Table 2 reports 24.6 mAP with (0,0) LiDAR FOV; the mechanism by which the LiDAR-based head operates without LiDAR features is not explained.
  • domain assumption Simulated limited FOV and point-drop corruptions are representative of real sensor malfunctions
    Robustness is only evaluated on synthetic corruptions; no real-world failure data is considered.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Reliability-Driven LiDAR-Camera Fusion for Robust 3D Object Detection." pith.science (2026). https://pith.science/paper/WF674QDP

@misc{pith2026250201856,
  author       = {Pith},
  title        = {Pith review of: Reliability-Driven LiDAR-Camera Fusion for Robust 3D Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WF674QDP}},
  note         = {Machine review of arXiv:2502.01856}
}
read the original abstract

Accurate and robust 3D object detection is essential for autonomous driving, where fusing data from sensors like LiDAR and camera enhances detection accuracy. However, sensor malfunctions such as corruption or disconnection can degrade performance, and existing fusion models often struggle to maintain reliability when one modality fails. To address this, we propose ReliFusion, a novel LiDAR-camera fusion framework operating in the bird's-eye view (BEV) space. ReliFusion integrates three key components: the Spatio-Temporal Feature Aggregation (STFA) module, which captures dependencies across frames to stabilize predictions over time; the Reliability module, which assigns confidence scores to quantify the dependability of each modality under challenging conditions; and the Confidence-Weighted Mutual Cross-Attention (CW-MCA) module, which dynamically balances information from LiDAR and camera modalities based on these confidence scores. Experiments on the nuScenes dataset show that ReliFusion significantly outperforms state-of-the-art methods, achieving superior robustness and accuracy in scenarios with limited LiDAR fields of view and severe sensor malfunctions.

Figures

Figures reproduced from arXiv: 2502.01856 by the authors.

Figure 1
Figure 1. Illustration of ReliFusion’s approach compared to previous [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overal architecture of ReliFusion. ReliFusion addresses the limitations of existing methods by introducing a Reliability module that dynamically adjusts the contribution of LiDAR and camera features based on real-time confidence scores. These scores, derived through CMCL, guide the CW-MCA module for robust multimodal fusion. Additionally, a STFA module enhances detection sta￾bility by leveraging cross-frame depe… view at source ↗
Figure 3
Figure 3. Qualitative detection results of BEVFusion and ReliFusion under LiDAR malfunctions scenarios. Clearly, BEVFusion struggles [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 25 canonical work pages

  1. [1]

    A survey of motion planning and control tech- niques for self-driving urban vehicles,

    B. Paden, M. ˇC´ap, S. Z. Yong, D. Yershov, and E. Fraz- zoli, “A survey of motion planning and control tech- niques for self-driving urban vehicles,” IEEE Transac- tions on intelligent vehicles , vol. 1, no. 1, pp. 33–55, 2016

    work page 2016

  2. [2]

    Pseudo-mono for monocular 3d object detection in au- tonomous driving,

    C. Tao, J. Cao, C. Wang, Z. Zhang, and Z. Gao, “Pseudo-mono for monocular 3d object detection in au- tonomous driving,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 33, no. 8, pp. 3962– 3975, 2023

    work page 2023

  3. [3]

    Pipc-3ddet: Har- nessing perspective information and proposal correla- tion for 3d point cloud object detection,

    C. Yu, B. Peng, Q. Huang, and J. Lei, “Pipc-3ddet: Har- nessing perspective information and proposal correla- tion for 3d point cloud object detection,” IEEE Trans- actions on Circuits and Systems for Video Technology , 2023

    work page 2023

  4. [4]

    Mvx-net: Mul- timodal voxelnet for 3d object detection,

    V . A. Sindagi, Y . Zhou, and O. Tuzel, “Mvx-net: Mul- timodal voxelnet for 3d object detection,” in 2019 In- ternational Conference on Robotics and Automation (ICRA). IEEE, 2019, pp. 7276–7282

    work page 2019

  5. [5]

    Pointaug- menting: Cross-modal augmentation for 3d object de- tection,

    C. Wang, C. Ma, M. Zhu, and X. Yang, “Pointaug- menting: Cross-modal augmentation for 3d object de- tection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 11 794–11 803

    work page 2021

  6. [6]

    Autoalign: Pixel-instance fea- ture aggregation for multi-modal 3d object detection,

    Z. Chen, Z. Li, S. Zhang, L. Fang, Q. Jiang, F. Zhao, B. Zhou, and H. Zhao, “Autoalign: Pixel-instance fea- ture aggregation for multi-modal 3d object detection,” arXiv preprint arXiv:2201.06493, 2022

    arXiv 2022

  7. [7]

    Bevfusion: A simple and robust lidar-camera fusion framework,

    T. Liang, H. Xie, K. Yu, Z. Xia, Z. Lin, Y . Wang, T. Tang, B. Wang, and Z. Tang, “Bevfusion: A simple and robust lidar-camera fusion framework,” Advances in Neural Information Processing Systems, vol. 35, pp. 10 421–10 434, 2022

    work page 2022

  8. [8]

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

    Z. Liu, H. Tang, A. Amini, X. Yang, H. Mao, D. L. Rus, and S. Han, “Bevfusion: Multi-task multi-sensor fusion with unified bird’s-eye view representation,” in 2023 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2023, pp. 2774–2781

    work page 2023

Show all 34 references
  1. [9]

    Benchmark- ing the robustness of lidar-camera fusion for 3d object detection,

    K. Yu, T. Tao, H. Xie, Z. Lin, T. Liang, B. Wang, P. Chen, D. Hao, Y . Wang, and X. Liang, “Benchmark- ing the robustness of lidar-camera fusion for 3d object detection,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 3187–3197

    2023

  2. [10]

    nuscenes: A multimodal dataset for autonomous driving,

    H. Caesar, V . Bankiti, A. H. Lang, S. V ora, V . E. Liong, Q. Xu, A. Krishnan, Y . Pan, G. Baldan, and O. Bei- jbom, “nuscenes: A multimodal dataset for autonomous driving,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2020, pp. 11 621–11 631

    2020

  3. [11]

    Pointnet++: Deep hierarchical feature learning on point sets in a met- ric space,

    C. R. Qi, L. Yi, H. Su, and L. J. Guibas, “Pointnet++: Deep hierarchical feature learning on point sets in a met- ric space,” Advances in neural information processing systems, vol. 30, 2017

    2017

  4. [12]

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

    Y . Zhou and O. Tuzel, “V oxelnet: End-to-end learning for point cloud based 3d object detection,” in Proceed- ings of the IEEE conference on computer vision and pat- tern recognition, 2018, pp. 4490–4499

    2018

  5. [13]

    Second: Sparsely embed- ded convolutional detection,

    Y . Yan, Y . Mao, and B. Li, “Second: Sparsely embed- ded convolutional detection,” Sensors, vol. 18, no. 10, p. 3337, 2018

    2018

  6. [14]

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

    S. Shi, C. Guo, L. Jiang, Z. Wang, J. Shi, X. Wang, and H. Li, “Pv-rcnn: Point-voxel feature set abstraction for 3d object detection,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 10 529–10 538

    2020

  7. [15]

    Detr3d: 3d object detection from multi- view images via 3d-to-2d queries,

    Y . Wanget al., “Detr3d: 3d object detection from multi- view images via 3d-to-2d queries,” inProceedings of the Conference on Robot Learning (CoRL), 2022

    2022

  8. [16]

    Bevdepth: Acquisition of reliable depth for multi-view 3d object detection,

    Y . Li, Z. Ge, G. Yu, J. Yang, Z. Wang, Y . Shi, J. Sun, and Z. Li, “Bevdepth: Acquisition of reliable depth for multi-view 3d object detection,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, no. 2, 2023, pp. 1477–1485

    2023

  9. [17]

    Msmdfusion: Fusing lidar and camera at mul- tiple scales with multi-depth seeds for 3d object de- tection,

    Y . Jiao, Z. Jie, S. Chen, J. Chen, L. Ma, and Y .-G. Jiang, “Msmdfusion: Fusing lidar and camera at mul- tiple scales with multi-depth seeds for 3d object de- tection,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2023, pp. 21 643–21 652

    2023

  10. [18]

    Gafusion: Adap- tive fusing lidar and camera with multiple guidance for 3d object detection,

    X. Li, B. Fan, J. Tian, and H. Fan, “Gafusion: Adap- tive fusing lidar and camera with multiple guidance for 3d object detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion, 2024, pp. 21 209–21 218

    2024

  11. [19]

    Lift, splat, shoot: Encoding im- ages from arbitrary camera rigs by implicitly unproject- ing to 3d,

    J. Philion and S. Fidler, “Lift, splat, shoot: Encoding im- ages from arbitrary camera rigs by implicitly unproject- ing to 3d,” in Computer Vision–ECCV 2020: 16th Eu- ropean Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XIV 16 . Springer, 2020, pp. 194– 210

    2020

  12. [20]

    Cross modal transformer: Towards fast and robust 3d object detection,

    J. Yan, Y . Liu, J. Sun, F. Jia, S. Li, T. Wang, and X. Zhang, “Cross modal transformer: Towards fast and robust 3d object detection,” in Proceedings of the IEEE/CVF International Conference on Computer Vi- sion, 2023, pp. 18 268–18 278

    2023

  13. [21]

    Bevdet4d: Exploit temporal cues in multi-camera 3d object detection,

    J. Huang et al. , “Bevdet4d: Exploit temporal cues in multi-camera 3d object detection,” arXiv preprint arXiv:2203.17054, 2022

    2022 arXiv

  14. [22]

    Bevformer: Learning bird’s-eye-view rep- resentation from multi-camera images via spatiotempo- ral transformers,

    Z. Li et al., “Bevformer: Learning bird’s-eye-view rep- resentation from multi-camera images via spatiotempo- ral transformers,” in Proceedings of the European Con- ference on Computer Vision (ECCV), 2022

    2022

  15. [23]

    3d-vid: Multi-frame point cloud feature fusion based on attention mechanisms for 3d object de- tection,

    Z. Zhai et al., “3d-vid: Multi-frame point cloud feature fusion based on attention mechanisms for 3d object de- tection,” Sensors, vol. 22, no. 22, p. 8640, 2022

    2022

  16. [24]

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

    X. Bai, Z. Hu, X. Zhu, Q. Huang, Y . Chen, H. Fu, and C.-L. Tai, “Transfusion: Robust lidar-camera fusion for 3d object detection with transformers,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 1090–1099

    2022

  17. [25]

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

    Y . Xieet al., “Sparsefusion: Fusing multi-modal sparse representations for multi-sensor 3d object detection,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023

    2023

  18. [26]

    Swin transformer: Hierarchical vision transformer using shifted windows,

    Z. Liu, Y . Lin, Y . Cao, H. Hu, Y . Wei, Z. Zhang, S. Lin, and B. Guo, “Swin transformer: Hierarchical vision transformer using shifted windows,” in Proceedings of the IEEE/CVF international conference on computer vi- sion, 2021, pp. 10 012–10 022

    2021

  19. [27]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” Advances in neural infor- mation processing systems, vol. 30, 2017

    2017

  20. [28]

    Is space- time attention all you need for video understanding?

    G. Bertasius, H. Wang, and L. Torresani, “Is space- time attention all you need for video understanding?” in ICML, vol. 2, no. 3, 2021, p. 4

    2021

  21. [29]

    Gaussian error linear units (gelus),

    D. Hendrycks and K. Gimpel, “Gaussian error linear units (gelus),” arXiv preprint arXiv:1606.08415, 2016

    2016 arXiv

  22. [30]

    Center-based 3d object detection and tracking,

    T. Yin, X. Zhou, and P. Krahenbuhl, “Center-based 3d object detection and tracking,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 11 784–11 793

    2021

  23. [31]

    Futr3d: A unified sensor fusion framework for 3d de- tection,

    X. Chen, T. Zhang, Y . Wang, Y . Wang, and H. Zhao, “Futr3d: A unified sensor fusion framework for 3d de- tection,” in proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023, pp. 172– 181

    2023

  24. [32]

    MMDetection: Open mmlab detection tool- box and benchmark,

    K. Chen, J. Wang, J. Pang, Y . Cao, Y . Xiong, X. Li, S. Sun, W. Feng, Z. Liu, J. Xu, Z. Zhang, D. Cheng, C. Zhu, T. Cheng, Q. Zhao, B. Li, X. Lu, R. Zhu, Y . Wu, J. Dai, J. Wang, J. Shi, W. Ouyang, C. C. Loy, and D. Lin, “MMDetection: Open mmlab detection tool- box and benchm...

    1906 arXiv

  25. [33]

    Transformer-based rgb and lidar fusion for enhanced object detection,

    R. Sadeghian, N. Hooshyaripour, and W. Lee, “Transformer-based rgb and lidar fusion for enhanced object detection,” in International Conference on Pat- tern Recognition. Springer, 2024, pp. 445–460

    2024

  26. [34]

    Patches are all you need?

    A. Trockman and J. Z. Kolter, “Patches are all you need?” arXiv preprint arXiv:2201.09792, 2022

    2022 arXiv

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.