Pith. sign in

REVIEW 3 major objections 7 minor 60 references

Revisiting Radar Camera Alignment by Contrastive Learning for 3D Object Detection

T0 review · 3 major / 7 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read RCAlign is a radar-camera 3D detector that aligns sparse queries across modalities with contrastive learning, reporting 67.3% NDS on nuScenes and a 4.3-point NDS gain over the prior fusion method.

desk verdict A credible radar-camera fusion paper whose dual-route contrastive alignment is a real architectural idea, though the alignment guarantee is overstated and the KD loss has a typo-level problem. read the letter →

arxiv 2504.16368 v1 pith:XBBOCTLN submitted 2025-04-23 cs.CV

Linhua Kong , Dongxia Chang , Lian Liu , Zisen Kong , Pengyuan Li , Yao Zhao This is my paper · ORCID
classification cs.CV
keywords radar-camerafusion3Dobjectdetectioncontrastivelearningdual-routealignmentknowledgedistillationbird's-eyeviewsparsequeriesnuScenes
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

The paper argues that the weak point of radar-camera 3D object detection is not the sensors but the way their features are aligned: previous methods either align dense bird's-eye-view grids without letting the two modalities interact, or use sparse queries as a pivot but never make the two modality views of the same query agree. RCAlign addresses this with a Dual-Route Alignment module in which every sparse query is updated twice, once by sampling radar features first and image features second, once in the opposite order, and a contrastive loss then pulls the two resulting features together when they carry the same index. A second module, Radar Feature Enhancement, densifies the sparse radar BEV features by distilling occupancy information derived from predicted 3D box centers. The authors report that this combination reaches 67.3% NDS on the nuScenes test set and 61.1% NDS / 53.7% mAP on the validation set, beating the previous radar-camera fusion method RCBEVDet by 4.3 NDS and 8.4 mAP.

What carries the argument

The load-bearing object is the Dual-Route Alignment (DRA) module. DRA takes a set of sparse queries (initial queries, temporal queries, and radar queries selected by the top-k peaks of a radar heatmap) and sends each query down two deformable-attention paths: radar-then-image and image-then-radar. Deformable attention lets each query sample a small set of offset locations around its reference point instead of attending globally. The two updated queries are normalized and matched by a contrastive loss (Eq. 4) whose target is the identity matrix $I$, so same-index pairs are pulled together and different-index pairs are pushed apart; the fused query is the element-wise sum. The second mechanism is Radar Feature Enhancement (RFE): predicted 3D box centers are projected onto BEV grids to form occupancy features, which are concatenated with radar BEV features and passed through a three-layer conv block to produce dense radar features; a knowledge-distillation loss (Eq. 5) then transfers these dense features back into the original radar branch, and a shared radar head supervises both.

What would settle it

Record the sampled deformable-attention locations for paired same-index queries during validation. If a substantial fraction of pairs attend to regions belonging to different ground-truth objects or to locations far apart, the index correspondence is broken and the contrastive loss is pushing the wrong pairs together. A cleaner experiment: replace the identity-matrix target with a Hungarian assignment between the two query sets based on predicted box centers; if performance does not drop, the index-based contrastive loss is not doing the alignment work attributed to it.

Watch

Extended reading notes

Core claim

RCAlign's central claim is that true cross-modal alignment for radar and camera is achievable by forcing two complementary routing orders to agree on the same sparse queries. The Dual-Route Alignment module produces a radar-first query and an image-first query for each object hypothesis, aligns them with a contrastive loss whose target matrix is the identity, and fuses them by element-wise addition. According to the paper this yields features that represent the same object at the same location in both modalities, which dense-BEV alignment and single-pivot sparse alignment do not guarantee. With the Radar Feature Enhancement module distilling denser radar features from predicted box centers, the method reports state-of-the-art 3D detection numbers on nuScenes, including 67.3% NDS on the test set and 61.1% NDS / 53.7% mAP on validation.

Load-bearing premise

The contrastive loss assumes that the two updated queries with the same index, one from the radar-first route and one from the image-first route, always describe the same physical object at the same location; if one route samples a different object, duplicates a query, or misses the object entirely, the loss pulls unrelated features together rather than aligning a matched pair.

Editorial extensions

If this is right

  • If the reported numbers hold, real-time radar-camera fusion can exceed 60% NDS on nuScenes, a threshold no previous radar-camera method in this comparison reached.
  • The ablation study places the source of the gain: adding radar features to the StreamPETR baseline brings the largest jump, while the contrastive alignment and the second deformable-attention step each contribute further, and the full DRA adds 6.6 NDS / 9.5 mAP over the camera-only baseline.
  • Velocity estimation improves sharply (mAVE 0.192 on validation vs 0.220 for RCBEVDet), consistent with radar's direct Doppler information surviving the fusion and alignment.
  • Per-class results show the biggest improvements on small and clustered objects such as pedestrians and motorcycles, suggesting the alignment mainly helps where radar returns are sparse.

Reading between the lines

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

  • An unstated risk: the contrastive loss assumes a one-to-one index correspondence between the two routes' queries. If duplicate queries or missed objects break that correspondence, the loss actively repels or attracts the wrong pairs; a permutation-invariant assignment step could be tested as a drop-in replacement.
  • The RFE module is circular in an interesting way: it densifies radar features using box centers predicted from the fusion queries, so early-training noise in the detector feeds back into the radar branch; starting the distillation only after the detector has stabilized is a natural extension.
  • Because DRA's queries are modality-agnostic, the same two-route alignment could be applied to lidar-camera fusion or radar-lidar-camera fusion without architectural change.
  • The robustness table suggests radar-camera fusion with densified radar can beat lidar-only detection in rain and at night; this invites tests on other adverse-weather datasets to check whether the nuScenes result generalizes.
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

3 major / 7 minor

Summary. The paper proposes RCAlign, a radar-camera fusion architecture for 3D object detection built on sparse query-based alignment. The key contributions are a Dual-Route Alignment (DRA) module that updates sparse queries through two deformable-attention paths (radar-then-image and image-then-radar) and aligns the resulting query pairs with a contrastive loss, and a Radar Feature Enhancement (RFE) module that densifies radar BEV features by distilling from features enhanced by predicted 3D box center occupancy. Experiments on nuScenes report state-of-the-art results: 67.3% NDS on the test set and 61.1% NDS / 53.7% mAP on the validation set with R50, surpassing RCBEVDet by 4.3% NDS and 8.4% mAP. Ablations decompose gains into DRA and RFE submodules, and robustness results are reported across weather and lighting conditions.

Significance. If the central claim holds, the paper demonstrates that dual-route cross-modal attention with contrastive query alignment is an effective fusion strategy for radar-camera 3D detection, and it provides a strong new state of the art on nuScenes. The manuscript includes extensive experiments: main tables, per-class AP, an ablation study isolating each component, a parametric analysis, robustness analysis, and tracking results. The benchmark claim is externally grounded on held-out nuScenes val/test splits. However, the paper does not report seed variance or release code, and the central alignment mechanism rests on an unverified index-correspondence assumption that the paper's own text partially contradicts.

major comments (3)
  1. [§III-C, Eq. (4); §III-D] The contrastive loss in Eq. (4) treats each query index as its own class through the identity target matrix I, pulling z_ri and zir of the same index together and pushing every different-index pair apart. This is only a valid alignment signal if query index is a proxy for object identity. The paper does not verify this premise with any matching or correspondence analysis, and §III-D explicitly states that 'the number of fused queries exceeds the count of ground truth boxes by a considerable margin' and that 'there will be cases where multiple queries predict the same box.' With duplicate queries, two different indices can represent the same physical object, yet Eq. (4) would actively repel their features across the two routes. The claimed 'guarantee' of same-object alignment in the introduction is therefore unsupported; the CL gain in Table V could stem from a generic feature-regularization effect rather than the stated mechanism. Please either reformulate the contrastive target to operate on matched object instances (e.g., using predicted boxes or Hungarian assignment), or provide an empirical analysis showing that same-index queries from the two routes correspond to the same ground-truth object in the majority of cases, and quantify the fraction of duplicate-query cases.
  2. [§III-D, Eq. (5)] As written, Eq. (5) defines LKD as the sum of (fer - fr) over all spatial and channel locations, with no absolute value or square. Since the sum of signed differences can be negative and arbitrarily large in magnitude, this is not a proper L1 or L2 distillation loss. If the intended loss is L1 or L2, the equation must include |·| or (·)^2; if the signed sum is intentional, its behavior as a training objective is unexplained and likely counterproductive because positive and negative errors could cancel. This issue directly affects the reported RFE gains, so the equation and the implemented loss need to be reconciled.
  3. [Table V and Tables VIII–IX] The paper reports no seed variance or confidence intervals for any of the results. This is particularly important for the core attribution claim: in Table V, the contrastive-loss ablation improves NDS from 0.588 to 0.592 (0.4%), and the full RFE addition improves NDS by 0.8%, while Tables VIII–IX show sensitivity to λ3 and λ4. Without multiple seeds or a stated evaluation protocol, these small differences may not be statistically distinguishable from training noise. Please report mean and standard deviation over at least three seeds for the key ablations, and state whether the reported main results are single runs.
minor comments (7)
  1. [General] The paper contains several typographical errors, including 'moudle' (§III-C), 'noval' (§III), 'Mutil-Modal' (Index Terms), 'Pre-class AP' (should be 'Per-class AP'), and 'Constrastive Loss Wight' (Table VIII heading).
  2. [§III-D, Eq. (5)] The subscript/superscript notation in Eq. (5) is inconsistent: the summation indices are H, W, C, but the term uses f_er^{ijk} and f_r^{ijk} without defining i, j, k. Please clarify the indexing.
  3. [Table V] The row/column structure of Table V is hard to parse: the meanings of the symbols in the first two rows ('C R DRA RH RFE' and 'DA SDA CL SRH KD') are not self-explanatory, and it is unclear which ablation corresponds to exactly which configuration. Adding explicit configuration names or checkmarks for each row would greatly improve readability.
  4. [§IV-A] The paper states that models are trained on NVIDIA 3090 GPUs but does not specify the number of GPUs, per-GPU batch size, or total training time. This information is needed for reproducibility, especially given the reported FPS numbers.
  5. [§IV-B] The claim that RCAlign 'enhances baseline (StreamPETR) results' is slightly misleading because the method includes a radar branch and additional modules; StreamPETR is a camera-only baseline, not an architectural baseline of the proposed method. Please clarify the baseline definition.
  6. [References] Reference [42] is cited for contrastive learning, but it is the CLIP paper; the standard contrastive learning references (e.g., SimCLR, MoCo) are also cited elsewhere. Using the CLIP citation for the contrastive loss definition is confusing and should be corrected.
  7. [General] The paper does not mention whether code or model weights will be released. Given the strong benchmark claims and the need to verify the contrastive-alignment mechanism, a code release or a detailed reproducibility appendix would strengthen the paper.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity identified; the reported gains rest on held-out nuScenes evaluation against external baselines.

full rationale

The paper's central claim, that RCAlign achieves state-of-the-art radar-camera 3D detection on nuScenes, is supported by measurements on the held-out validation and test splits (Tables I and III) and by ablations (Table V) that compare architectural variants on the same external metric. The DRA contrastive loss (Eq. 4) is a training objective built on an identity target over query indices; it does not by construction determine the reported NDS or mAP, which are produced by the downstream 3D detection head and evaluated against ground-truth boxes. The RFE module uses the model's own predicted 3D box centers to form occupancy features and then applies a knowledge-distillation loss (Eq. 5); this is a self-distillation/self-training regularizer, not a fitted parameter that is later renamed as a prediction. The reference list contains no prior works by the present authors, and the compared methods (RCBEVDet, CRN, StreamPETR, etc.) are external, so there is no load-bearing self-citation chain. One internal tension exists: Eq. 4 treats each query index as its own class, while Section III-D acknowledges that multiple fused queries can predict the same box. That tension threatens the semantic claim that same-index queries always correspond to the same physical object, but it is a correctness or robustness concern about the alignment mechanism, not a circular derivation. Even if the contrastive target were mis-specified, the final detection numbers would not be equivalent to the loss construction by definition. Hence the derivation chain is not circular.

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

The paper's central claim rests on the external benchmark being a valid measure, on query index identity standing for object identity in the contrastive loss, on constant-velocity radar point compensation, and on the usefulness of self-derived occupancy features as radar supervision. Loss weights and the number of radar queries are chosen on the validation set and should be viewed as fitted hyperparameters, not first principles.

free parameters (4)
  • Contrastive loss weight lambda3 = 1
    Chosen after sweeping 0.1, 1, 10 on nuScenes val (Table VIII).
  • Knowledge distillation loss weight lambda4 = 5
    Chosen after sweeping 1, 5, 10 on nuScenes val (Table IX).
  • Radar head auxiliary loss weights lambda1 and lambda2 = 1, 1
    Set to 1 in Section IV-A; no sensitivity analysis is reported.
  • Number of top-k radar queries = 30
    Set in Section IV-A; the topk sampling rule is selected over random and FPS in Table VI.
assumptions (4)
  • domain assumption nuScenes annotations and official evaluation metrics are a valid measure of 3D detection quality for the SOTA claim.
    All comparisons and conclusions in Section IV are drawn from the nuScenes val and test metrics.
  • domain assumption Sparse queries sharing an index across the two routing paths represent the same physical object, making the identity matrix a valid contrastive target.
    Eq. 4 creates positive pairs by index; if the correspondence fails, contrastive learning would pull different objects together.
  • domain assumption Five-frame radar aggregation with constant-velocity compensation approximates object motion well enough for BEV feature fusion.
    Eq. 1 shifts past radar points using velocity times time; this ignores acceleration and ego-motion corrections not described.
  • ad hoc to paper Predicted 3D box centers carry sufficient signal to supervise radar BEV densification.
    The RFE module in Section III-D uses the model's own box predictions as occupancy supervision; if early predictions are biased, self-distillation can reinforce errors.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Revisiting Radar Camera Alignment by Contrastive Learning for 3D Object Detection." pith.science (2026). https://pith.science/paper/XBBOCTLN

@misc{pith2026250416368,
  author       = {Pith},
  title        = {Pith review of: Revisiting Radar Camera Alignment by Contrastive Learning for 3D Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XBBOCTLN}},
  note         = {Machine review of arXiv:2504.16368}
}
read the original abstract

Recently, 3D object detection algorithms based on radar and camera fusion have shown excellent performance, setting the stage for their application in autonomous driving perception tasks. Existing methods have focused on dealing with feature misalignment caused by the domain gap between radar and camera. However, existing methods either neglect inter-modal features interaction during alignment or fail to effectively align features at the same spatial location across modalities. To alleviate the above problems, we propose a new alignment model called Radar Camera Alignment (RCAlign). Specifically, we design a Dual-Route Alignment (DRA) module based on contrastive learning to align and fuse the features between radar and camera. Moreover, considering the sparsity of radar BEV features, a Radar Feature Enhancement (RFE) module is proposed to improve the densification of radar BEV features with the knowledge distillation loss. Experiments show RCAlign achieves a new state-of-the-art on the public nuScenes benchmark in radar camera fusion for 3D Object Detection. Furthermore, the RCAlign achieves a significant performance gain (4.3\% NDS and 8.4\% mAP) in real-time 3D detection compared to the latest state-of-the-art method (RCBEVDet).

Figures

Figures reproduced from arXiv: 2504.16368 by the authors.

Figure 1
Figure 1. Different alignment methods: Dense BEV Alignment (a), Sparse BEV Alignment (b) and Dual-Route Alignment (ours c). Q, V, BEV, PV and CL [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The overall architecture of RCAlign. Multi-view images and radar points are fed into the backbone to extract modal-specific features. Then radar [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The proposed Dual-Route Alignment (DRA) module (a) and Radar Feature Enhancement (RFE) module (b). The DRA firstly utilizes sparse queries [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Visualisation results of RCAlign. The red and blue boxes indicate ground truth and prediction, respectively. The orange and green circles indicate [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: More visualisation results. The first two rows indicate day, the middle two rows indicate night, and the last two rows indicate rain. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 21 canonical work pages

  1. [1]

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

    J. Huang, G. Huang, Z. Zhu, Y . Ye, and D. Du, “Bevdet: High- performance multi-camera 3d object detection in bird-eye-view,” arXiv preprint arXiv:2112.11790, 2021

    arXiv 2021

  2. [2]

    Petr: Position embedding trans- formation for multi-view 3d object detection,

    Y . Liu, T. Wang, X. Zhang, and J. Sun, “Petr: Position embedding trans- formation for multi-view 3d object detection,” in European Conference on Computer Vision . Springer, 2022, pp. 531–548

    2022

  3. [3]

    Exploring object- centric temporal modeling for efficient multi-view 3d object detection,

    S. Wang, Y . Liu, T. Wang, Y . Li, and X. Zhang, “Exploring object- centric temporal modeling for efficient multi-view 3d object detection,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 3621–3631

    2023

  4. [4]

    Bevnext: Reviving dense bev frameworks for 3d object detection,

    Z. Li, S. Lan, J. M. Alvarez, and Z. Wu, “Bevnext: Reviving dense bev frameworks for 3d object detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 20 113–20 123

    2024

  5. [5]

    3dppe: 3d point positional encoding for transformer-based multi-camera 3d object detection,

    C. Shu, J. Deng, F. Yu, and Y . Liu, “3dppe: 3d point positional encoding for transformer-based multi-camera 3d object detection,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 3580–3589

    work page 2023

  6. [6]

    Graph-detr4d: Spatio-temporal graph modeling for multi-view 3d object detection,

    Z. Chen, Z. Chen, Z. Li, S. Zhang, L. Fang, Q. Jiang, F. Wu, and F. Zhao, “Graph-detr4d: Spatio-temporal graph modeling for multi-view 3d object detection,” IEEE Transactions on Image Processing , 2024

    work page 2024

  7. [7]

    Unleashing hydra: Hybrid fusion, depth consistency and radar for unified 3d perception,

    P. Wolters, J. Gilg, T. Teepe, F. Herzog, A. Laouichi, M. Hofmann, and G. Rigoll, “Unleashing hydra: Hybrid fusion, depth consistency and radar for unified 3d perception,” arXiv preprint arXiv:2403.07746, 2024

    arXiv 2024

  8. [8]

    Rcbevdet: Radar-camera fusion in bird’s eye view for 3d object detection,

    Z. Lin, Z. Liu, Z. Xia, X. Wang, Y . Wang, S. Qi, Y . Dong, N. Dong, L. Zhang, and C. Zhu, “Rcbevdet: Radar-camera fusion in bird’s eye view for 3d object detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 14 928–14 937. JOURNAL OF LATEX CLASS FILES, VOL. 18, NO. 9, SEPTEMBER 2020 11

    work page 2024

Show all 60 references
  1. [9]

    Futr3d: A unified sensor fusion framework for 3d detection,

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

    2023

  2. [10]

    Sparsefusion3d: Sparse sensor fusion for 3d object detection by radar and camera in environmental perception,

    Z. Yu, W. Wan, M. Ren, X. Zheng, and Z. Fang, “Sparsefusion3d: Sparse sensor fusion for 3d object detection by radar and camera in environmental perception,” IEEE Transactions on Intelligent Vehicles , 2023

    2023

  3. [11]

    Crn: Camera radar net for accurate, robust, efficient 3d perception,

    Y . Kim, J. Shin, S. Kim, I.-J. Lee, J. W. Choi, and D. Kum, “Crn: Camera radar net for accurate, robust, efficient 3d perception,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023, pp. 17 615–17 626

    2023

  4. [12]

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

    X. Zhu, W. Su, L. Lu, B. Li, X. Wang, and J. Dai, “Deformable detr: Deformable transformers for end-to-end object detection,”arXiv preprint arXiv:2010.04159, 2020

    2010 arXiv

  5. [13]

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

    T. Wang, X. Zhu, J. Pang, and D. Lin, “Fcos3d: Fully convolutional one- stage monocular 3d object detection,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021, pp. 913–922

    2021

  6. [14]

    Objects as points,

    X. Zhou, D. Wang, and P. Kr ¨ahenb¨uhl, “Objects as points,” arXiv preprint arXiv:1904.07850, 2019

    1904 arXiv

  7. [15]

    Progressive coordinate transforms for monocular 3d object detection,

    L. Wang, L. Zhang, Y . Zhu, Z. Zhang, T. He, M. Li, and X. Xue, “Progressive coordinate transforms for monocular 3d object detection,” Advances in Neural Information Processing Systems , vol. 34, pp. 13 364–13 377, 2021

    2021

  8. [16]

    Stereoscopic vision recalling memory for monocular 3d object detection,

    J. U. Kim, H.-I. Kim, and Y . M. Ro, “Stereoscopic vision recalling memory for monocular 3d object detection,” IEEE Transactions on Image Processing, vol. 32, pp. 2749–2760, 2023

    2023

  9. [17]

    Obmo: One bounding box multiple objects for monocular 3d object detection,

    C. Huang, T. He, H. Ren, W. Wang, B. Lin, and D. Cai, “Obmo: One bounding box multiple objects for monocular 3d object detection,” IEEE Transactions on Image Processing , vol. 32, pp. 6570–6581, 2023

    2023

  10. [18]

    Query-based temporal fusion with explicit motion for 3d object detection,

    J. Hou, Z. Liu, Z. Zou, X. Ye, X. Bai et al. , “Query-based temporal fusion with explicit motion for 3d object detection,” Advances in Neural Information Processing Systems , vol. 36, 2024

    2024

  11. [19]

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

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

    2022

  12. [20]

    Bevformer: Learning bird’s-eye-view representation from multi-camera images via spatiotemporal transformers,

    Z. Li, W. Wang, H. Li, E. Xie, C. Sima, T. Lu, Y . Qiao, and J. Dai, “Bevformer: Learning bird’s-eye-view representation from multi-camera images via spatiotemporal transformers,” in European conference on computer vision. Springer, 2022, pp. 1–18

    2022

  13. [21]

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

    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 detec- tion,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 37, no. 2, 2023, pp. 1477–1485

    2023

  14. [22]

    Bevformer v2: Adapting modern image backbones to bird’s-eye-view recognition via perspective supervision,

    C. Yang, Y . Chen, H. Tian, C. Tao, X. Zhu, Z. Zhang, G. Huang, H. Li, Y . Qiao, L. Lu et al. , “Bevformer v2: Adapting modern image backbones to bird’s-eye-view recognition via perspective supervision,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern ...

    2023

  15. [23]

    M 2 bev: Multi-camera joint 3d detection and segmentation with unified birds-eye view representation,

    E. Xie, Z. Yu, D. Zhou, J. Philion, A. Anandkumar, S. Fidler, P. Luo, and J. M. Alvarez, “M 2 bev: Multi-camera joint 3d detection and segmentation with unified birds-eye view representation,” arXiv preprint arXiv:2204.05088, 2022

    2022 arXiv

  16. [24]

    Bevstereo: Enhancing depth estimation in multi-view 3d object detection with temporal stereo,

    Y . Li, H. Bao, Z. Ge, J. Yang, J. Sun, and Z. Li, “Bevstereo: Enhancing depth estimation in multi-view 3d object detection with temporal stereo,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, no. 2, 2023, pp. 1486–1494

    2023

  17. [25]

    Cobev: Elevating roadside 3d object detection with depth and height complementarity,

    H. Shi, C. Pang, J. Zhang, K. Yang, Y . Wu, H. Ni, Y . Lin, R. Stiefel- hagen, and K. Wang, “Cobev: Elevating roadside 3d object detection with depth and height complementarity,” IEEE Transactions on Image Processing, 2024

    2024

  18. [26]

    Sparse4d: Multi-view 3d object detection with sparse spatial-temporal fusion,

    X. Lin, T. Lin, Z. Pei, L. Huang, and Z. Su, “Sparse4d: Multi-view 3d object detection with sparse spatial-temporal fusion,” arXiv preprint arXiv:2211.10581, 2022

    2022 arXiv

  19. [27]

    Lift, splat, shoot: Encoding images from arbitrary camera rigs by implicitly unprojecting to 3d,

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

    2020

  20. [28]

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

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

    2022 arXiv

  21. [29]

    Bevpoolv2: A cutting-edge implementation of bevdet toward deployment,

    ——, “Bevpoolv2: A cutting-edge implementation of bevdet toward deployment,” arXiv preprint arXiv:2211.17111 , 2022

    2022 arXiv

  22. [30]

    Centerfusion: Center-based radar and camera fusion for 3d object detection,

    R. Nabati and H. Qi, “Centerfusion: Center-based radar and camera fusion for 3d object detection,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision , 2021, pp. 1527–1536

    2021

  23. [31]

    Radiant: Radar-image association network for 3d object detection,

    Y . Long, A. Kumar, D. Morris, X. Liu, M. Castro, and P. Chakravarty, “Radiant: Radar-image association network for 3d object detection,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 37, no. 2, 2023, pp. 1808–1816

    2023

  24. [32]

    Craft: Camera-radar 3d object detection with spatio-contextual fusion transformer,

    Y . Kim, S. Kim, J. W. Choi, and D. Kum, “Craft: Camera-radar 3d object detection with spatio-contextual fusion transformer,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 37, no. 1, 2023, pp. 1160–1168

    2023

  25. [33]

    Mvfusion: Multi-view 3d object detection with semantic-aligned radar and camera fusion,

    Z. Wu, G. Chen, Y . Gan, L. Wang, and J. Pu, “Mvfusion: Multi-view 3d object detection with semantic-aligned radar and camera fusion,” in 2023 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2023, pp. 2766–2773

    2023

  26. [34]

    Dimensionality reduction by learning an invariant mapping,

    R. Hadsell, S. Chopra, and Y . LeCun, “Dimensionality reduction by learning an invariant mapping,” in 2006 IEEE computer society con- ference on computer vision and pattern recognition (CVPR’06) , vol. 2. IEEE, 2006, pp. 1735–1742

    2006

  27. [35]

    Momentum contrast for unsupervised visual representation learning,

    K. He, H. Fan, Y . Wu, S. Xie, and R. Girshick, “Momentum contrast for unsupervised visual representation learning,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 9729–9738

    2020

  28. [36]

    A simple framework for contrastive learning of visual representations,

    T. Chen, S. Kornblith, M. Norouzi, and G. Hinton, “A simple framework for contrastive learning of visual representations,” in International conference on machine learning . PMLR, 2020, pp. 1597–1607

    2020

  29. [37]

    Aligning pretraining for detection via object-level contrastive learning,

    F. Wei, Y . Gao, Z. Wu, H. Hu, and S. Lin, “Aligning pretraining for detection via object-level contrastive learning,” Advances in Neural Information Processing Systems , vol. 34, pp. 22 682–22 694, 2021

    2021

  30. [38]

    Fsce: Few-shot object detection via contrastive proposal encoding,

    B. Sun, B. Li, S. Cai, Y . Yuan, and C. Zhang, “Fsce: Few-shot object detection via contrastive proposal encoding,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2021, pp. 7352–7362

    2021

  31. [39]

    Cat-det: Contrastively augmented transformer for multi-modal 3d object detection,

    Y . Zhang, J. Chen, and D. Huang, “Cat-det: Contrastively augmented transformer for multi-modal 3d object detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 908–917

    2022

  32. [40]

    Feature pyramid networks for object detection,

    T.-Y . Lin, P. Doll´ar, R. Girshick, K. He, B. Hariharan, and S. Belongie, “Feature pyramid networks for object detection,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2017, pp. 2117–2125

    2017

  33. [41]

    Pointpillars: Fast encoders for object detection from point clouds,

    A. H. Lang, S. V ora, H. Caesar, L. Zhou, J. Yang, and O. Beijbom, “Pointpillars: Fast encoders for object detection from point clouds,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 12 697–12 705

    2019

  34. [42]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark et al., “Learning transferable visual models from natural language supervision,” in International conference on machine learning . PMLR, 2021, pp. 8748–8763

    2021

  35. [43]

    Batch normalization: Accelerating deep network training by reducing internal covariate shift,

    S. Ioffe and C. Szegedy, “Batch normalization: Accelerating deep network training by reducing internal covariate shift,” in International conference on machine learning . pmlr, 2015, pp. 448–456

    2015

  36. [44]

    Rectified linear units improve restricted boltz- mann machines,

    V . Nair and G. E. Hinton, “Rectified linear units improve restricted boltz- mann machines,” in Proceedings of the 27th international conference on machine learning (ICML-10) , 2010, pp. 807–814

    2010

  37. [45]

    Distilling the knowledge in a neural network,

    G. Hinton, O. Vinyals, and J. Dean, “Distilling the knowledge in a neural network,” arXiv preprint arXiv:1503.02531 , 2015

    2015 arXiv

  38. [46]

    Time will tell: New outlooks and a baseline for temporal multi-view 3d object detection,

    J. Park, C. Xu, S. Yang, K. Keutzer, K. M. Kitani, M. Tomizuka, and W. Zhan, “Time will tell: New outlooks and a baseline for temporal multi-view 3d object detection,” in The Eleventh International Confer- ence on Learning Representations , 2022

    2022

  39. [47]

    Sparsebev: High- performance sparse 3d object detection from multi-camera videos,

    H. Liu, Y . Teng, T. Lu, H. Wang, and L. Wang, “Sparsebev: High- performance sparse 3d object detection from multi-camera videos,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 18 580–18 590

    2023

  40. [48]

    Bridging the view disparity between radar and camera features for multi-modal fusion 3d object detection,

    T. Zhou, J. Chen, Y . Shi, K. Jiang, M. Yang, and D. Yang, “Bridging the view disparity between radar and camera features for multi-modal fusion 3d object detection,” IEEE Transactions on Intelligent Vehicles , vol. 8, no. 2, pp. 1523–1535, 2023

    2023

  41. [49]

    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. Beijbom, “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

  42. [50]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2016, pp. 770–778

    2016

  43. [51]

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

    Y . Lee, J.-w. Hwang, S. Lee, Y . Bae, and J. Park, “An energy and gpu- computation efficient backbone network for real-time object detection,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops , 2019, pp. 0–0. JOURNAL OF LATEX CLAS...

    2019

  44. [52]

    Imagenet: A large-scale hierarchical image database,

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, “Imagenet: A large-scale hierarchical image database,” in 2009 IEEE conference on computer vision and pattern recognition . Ieee, 2009, pp. 248–255

    2009

  45. [53]

    Is pseudo- lidar needed for monocular 3d object detection?

    D. Park, R. Ambrus, V . Guizilini, J. Li, and A. Gaidon, “Is pseudo- lidar needed for monocular 3d object detection?” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021, pp. 3142–3152

    2021

  46. [54]

    Focal-petr: Embracing foreground for efficient multi-camera 3d object detection,

    S. Wang, X. Jiang, and Y . Li, “Focal-petr: Embracing foreground for efficient multi-camera 3d object detection,” IEEE Transactions on Intelligent Vehicles, 2023

    2023

  47. [55]

    Dn-detr: Accelerate detr training by introducing query denoising,

    F. Li, H. Zhang, S. Liu, J. Guo, L. M. Ni, and L. Zhang, “Dn-detr: Accelerate detr training by introducing query denoising,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 13 619–13 627

    2022

  48. [56]

    Class-balanced grouping and sampling for point cloud 3d object detection,

    B. Zhu, Z. Jiang, X. Zhou, Z. Li, and G. Yu, “Class-balanced grouping and sampling for point cloud 3d object detection,” arXiv preprint arXiv:1908.09492, 2019

    1908 arXiv

  49. [57]

    Decoupled weight decay regularization,

    I. Loshchilov and F. Hutter, “Decoupled weight decay regularization,” arXiv preprint arXiv:1711.05101 , 2017

    2017 arXiv

  50. [58]

    Petrv2: A unified framework for 3d perception from multi-camera images,

    Y . Liu, J. Yan, F. Jia, S. Li, A. Gao, T. Wang, and X. Zhang, “Petrv2: A unified framework for 3d perception from multi-camera images,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 3262–3272

    2023

  51. [59]

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

    Y . Li, Y . Chen, X. Qi, Z. Li, J. Sun, and J. Jia, “Unifying voxel- based representation with transformer for 3d object detection,” Advances in Neural Information Processing Systems , vol. 35, pp. 18 442–18 455, 2022

    2022

  52. [60]

    Pointnet++: Deep hierarchical feature learning on point sets in a metric space,

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

    2017

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.