Pith. sign in

REVIEW 3 major objections 7 minor 54 references

HGSFusion: Radar-Camera Fusion with Hybrid Generation and Synchronization for 3D Object Detection

T0 review · 3 major / 7 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read HGSFusion claims that mask-guided hybrid point generation and dual radar-camera sync improve 3D detection, beating state-of-the-art by 6.53% RoI AP and 2.03% BEV AP.

desk verdict Solid radar-camera fusion paper with a real engineering contribution; the undocumented mask dependency and missing error bars are the main soft spots. read the letter →

arxiv 2412.11489 v1 pith:Y654EP66 submitted 2024-12-16 cs.CV cs.AIcs.LGcs.RO

Zijian Gu , Jianwei Ma , Yan Huang , Honghao Wei , Zhanye Chen , Hui Zhang , Wei Hong This is my paper · ORCID
classification cs.CVcs.AIcs.LGcs.RO
keywords radar-camerafusion3Dobjectdetectionbird'seyeviewpointcloudgenerationDOAestimationerror4Dimagingradarmulti-modalautonomousdriving
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

Millimeter-wave radar sees in all weather but returns sparse point clouds with angular errors, while cameras supply rich semantics but no reliable depth and degrade in bad light. HGSFusion claims both problems can be attacked in one network: generate extra radar points inside image-derived object masks, sampling near real foreground radar returns with a Gaussian and elsewhere uniformly, then fuse radar and image features in bird's-eye view with a two-step synchronization. On the VoD and TJ4DRadSet benchmarks the paper reports state-of-the-art results, beating prior fusion methods by 6.53 percentage points in driving-corridor AP and 2.03 percentage points in BEV AP. If the finding holds, it would make low-cost radar-camera systems substantially more competitive for autonomous perception.

What carries the argument

The machinery is a pair of modules. The Radar Hybrid Generation Module (RHGM) takes raw radar points projected onto the camera image, keeps those inside predicted instance masks as foreground points, builds a hybrid probability density function (Gaussian around each foreground point, uniform elsewhere inside the mask), samples hundreds of new points per mask, copies depth and physical features from the nearest foreground point, and encodes raw, foreground, and generated points with separate feature slots plus a one-hot point-type label so pillar pooling does not blur their identities. The Dual Sync Module (DSM) first multiplies image bird's-eye-view features by a radar-predicted spatial pattern of object presence (spatial sync), then concatenates radar and enhanced image features and predicts per-channel weights to combine them (modality sync). The hybrid points feed a radar pillar backbone, and the two sync steps produce the fused bird's-eye-view features that go into the detection head.

What would settle it

Corrupt the segmentation masks on VoD in a controlled way, for example randomly erode or drop one in five instance masks or shift masks by a few pixels, and rerun the full HGSFusion pipeline; if RoI AP does not drop substantially, then the reported gains are not actually carried by mask-guided generation. A complementary check is to compare the depth assigned to generated points against LiDAR ground truth on the same scene and measure whether depth errors concentrate on object boundaries.

Watch

Extended reading notes

Core claim

The paper's central claim is that radar point clouds should be treated as samples from a physically motivated probability distribution, not as fixed inputs. Because conventional radar processing estimates angles from beamforming peaks, the returned points can be systematically displaced; HGSFusion therefore generates a denser hybrid cloud by sampling a Gaussian centered on each foreground radar point that lands inside an image-instance mask, plus uniform samples in mask regions away from those points, and assigns each generated point the depth and radar features of its nearest foreground point. The paper further claims that fusing the two modalities in bird's-eye-view space requires synchronization: a radar-derived spatial pattern sharpens image features by highlighting where objects are, and a learned per-channel weighting keeps bad image features, for example in dark or shiny scenes, from corrupting the fused representation. On its own terms, the result is that this generation-plus-synchronization design outperforms existing radar-camera detectors on both benchmarks, with ablation studies attributing the gain to both modules and to a separate encoding scheme that preserves point identity inside pillars.

Load-bearing premise

The whole generation chain assumes the instance masks from the segmentation network are accurate and that assigning each synthetic point the depth and features of the single nearest foreground radar point gives correct physical values; when a mask is wrong or the nearest point lies on another object, the denser cloud is confidently wrong in ways that propagate into detection.

Editorial extensions

If this is right

  • Denser radar input generated from semantic masks improves car and pedestrian detection most, while cyclist detection can decline because similar objects such as parked bicycles and scooter racks pollute the masks, a drop the paper explicitly reports.
  • Spatial sync with radar positional features gives the largest robustness gain in dark and shiny lighting, where naive radar-camera fusion can hurt performance because image features are unreliable.
  • Separate encoding with distributed feature slots and point-type markers outperforms simple concatenation, indicating that pillar-based fusion needs explicit identity preservation to avoid mixing generated, foreground, and raw points.
  • Pure Gaussian or pure uniform generation both underperform the hybrid scheme, so the mixture ratio of Gaussian to uniform samples is a real design choice rather than a detail.
  • The complete system improves over its own radar-camera baseline by 4.14% EAA AP and 6.19% RoI AP on the VoD validation set, with both RHGM and DSM contributing separately.

Reading between the lines

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

  • The same mask-guided densification idea could transfer to LiDAR or to other sparse sensor pairs, but the depth-copying step would need to be replaced by an estimated depth prior to avoid propagating cross-object depth.
  • Because each generated point inherits depth and features from the single nearest foreground radar point, object boundaries are the likely weak point; an uncertainty-weighted or interpolation-based depth assignment is a natural extension the paper does not explore.
  • The method's dependence on a pretrained segmentation network is a testable liability: ablating mask confidence thresholds or measuring performance under mask noise would quantify how much of the gain is actually carried by segmentation quality.
  • The paper reports lighting-condition splits but not weather-condition splits; fog and rain are where radar's all-weather advantage should be largest, so a weather-annotated evaluation would be a direct stress test of the claimed robustness.
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 HGSFusion, a radar-camera fusion network for 3D object detection. The Radar Hybrid Generation Module (RHGM) densifies raw radar points by sampling additional points inside image instance masks according to a hybrid Gaussian/uniform distribution, assigning each generated point the depth and physical features of the nearest foreground radar point. The Dual Sync Module (DSM) first multiplies image BEV features with a radar-derived spatial presence pattern and then predicts channel-wise importance weights for modality fusion. Experiments on the View-of-Delft (VoD) and TJ4DRadSet datasets report state-of-the-art results, e.g., 79.46% RoI AP on VoD versus 72.93% for LXL, and 43.23% BEV mAP on TJ4DRadSet versus 41.20% for LXL. Ablations show contributions from RHGM, DSM, the hybrid generation scheme, and the separate point encoding strategy.

Significance. If the reported results are reliable, HGSFusion advances radar-camera fusion by explicitly modeling DOA estimation error through point generation and by using radar spatial patterns to guide image features. The paper ships a complete architecture, public-benchmark comparisons, ablations of each component, encoding-strategy comparisons, lighting-condition analysis, and code. These are concrete strengths. However, the significance is currently tempered by three reproducibility and correctness concerns: the segmentation network that supplies the instance masks is not described; generated points can inherit depth from a different object due to the unrestricted nearest-foreground-point assignment; and all quantitative claims rest on single runs without variance estimates. These issues are load-bearing for the central 'outperforms state-of-the-art' claim and need to be addressed before the results can be fully credited.

major comments (3)
  1. [RHGM, 'Obtaining foreground points'; Implementation Details] The manuscript does not specify how the Mask2former segmentation network is trained, which dataset(s) it was trained on, whether it was fine-tuned on VoD/TJ4DRadSet, or what its accuracy is. This is load-bearing because every foreground point and therefore every generated point is defined by these masks; if the masks are inaccurate or were trained on the benchmark's test distribution, the reported improvements could be an artifact of mask quality rather than of the proposed generation scheme. The authors' own appendix 'Explanation of the Cyclist Category in the VoD Dataset' concedes that mask quality directly causes a performance decline. Please provide the segmentation training protocol, a quantitative mask-quality measure (e.g., IoU or point-in-mask precision/recall), and, if feasible, an ablation that replaces predicted masks with ground-truth instance masks.
  2. [RHGM, 'Generating hybrid points'] Each generated point G_i is assigned the depth and physical features of the nearest foreground point P_fore, with no restriction that P_fore belongs to the same instance mask. If a mask contains no raw radar points, or a point from a neighboring object is globally nearest, the generated point receives wrong depth and features, is unprojected into 3D, and propagates into the radar BEV features. The authors do not analyze how often cross-mask assignment occurs or its effect on detection. I request either (a) restricting the nearest-neighbor search to foreground points within the same mask, or (b) providing an empirical analysis of cross-mask depth assignment and an ablation showing that the unrestricted assignment is not degrading or is negligible. This is a central mechanism of RHGM and directly affects the validity of the claimed gains.
  3. [Experiments, Tables 1-7] All results are reported from single training runs, with no error bars, multiple seeds, or statistical significance tests. The headline margins (e.g., 6.53% RoI AP over LXL on VoD, 2.03% BEV mAP on TJ4DRadSet) are plausible but may be within run-to-run variation for this type of detector, especially given the small validation/test sets. Please report mean and standard deviation over at least three seeds, or provide a significance test, for at least the main comparison tables and the ablation tables. Without this, the 'outperforms state-of-the-art' claim is not statistically grounded.
minor comments (7)
  1. [Throughout] The text uses 'lightning' where 'lighting' is meant (e.g., 'Influences of Lightning Conditions', 'adverse lightning'); please correct.
  2. [Throughout] The dataset name is inconsistently written as 'V oD' and 'VoD'; please standardize.
  3. [Eq. (3) and Implementation Details] The Gaussian PDF in Eq. (3) introduces standard deviations b1 and b2, but the hyperparameter details in the implementation section only list r and the number of points per mask. Please provide the values of b1 and b2 and state how they were selected.
  4. [Figures 2 and 8] The color legend is inconsistent: Figure 2 says generated, foreground, and raw points are green, orange, and blue, respectively, while Figure 8's caption says generated points are shown in orange. Please make the color coding uniform across figures.
  5. [Implementation Details] The segmentation network is referred to as 'Mask2former' but the cited work is 'Mask2Former'; please align the name with the reference.
  6. [Eq. (2) and Eq. (4)] The notation R_i(u,v) defines a set of pixel coordinates, but Eq. (4) treats it as a condition; the complement notation ∂Rm R_i is also unclear. Please rewrite these equations with clearer set notation, e.g., defining R_i as a subset of the mask region and using indicator functions.
  7. [Section 'Separate Radar Point Encoding'] The name 'Differentiable Encoding' is misleading; the one-hot point-type encoding is not different in differentiability from the other strategies. Consider renaming it e.g., 'Type Encoding' to avoid confusion.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central claim is an empirical benchmark result evaluated against held-out annotations, and no equation reduces the reported AP to the method's inputs.

full rationale

The paper's load-bearing claim is empirical: HGSFusion achieves higher RoI AP and BEV AP than prior methods on the VoD validation set and TJ4DRadSet test set using the official evaluation protocols. The Radar Hybrid Generation Module (RHGM) constructs denser radar points through Equations (1)-(4) and assigns depth/features from the nearest foreground point, but the reported detection AP is not defined in terms of this generation process; it is computed by matching predicted 3D boxes to ground-truth bounding boxes at fixed IoU thresholds. The Dual Sync Module's Equations (5)-(9) describe feature weighting and spatial enhancement, but again the final metric is externally evaluated on benchmark annotations, not on any quantity these equations are fitted to reproduce. Hyperparameter choices, such as the 50/200 Gaussian/uniform split, are studied as ablations (Figure 6 and Table 3) rather than fitted to the reported test numbers. The dependence on Mask2former instance masks and RadarPillarNet as an encoder introduces a potential correctness risk if mask quality is poor, but that is a component-quality concern, not a circular derivation: no step asserts that the generated points must receive correct depth by construction, nor does any equation define the benchmark AP as a function of the masks. There is also no load-bearing self-citation chain: the cited methods are baselines, backbones, or external published components, and no uniqueness theorem is imported to force the design. Overall, the derivation chain is self-contained with respect to the empirical claim, so circularity is not present.

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

The method relies on a semantic segmentation model and a set of data-generation hyperparameters rather than on new physical entities. The main load-bearing choices are the hand-picked generation counts and the Gaussian sigma values, one of which is not reported.

free parameters (3)
  • r (neighborhood radius in pixels) = 51
    Diameter of the area around each foreground point where Gaussian sampling is used. Chosen by hand; no sensitivity analysis for this value reported.
  • Points generated per mask = 250 (50 Gaussian, 200 uniform)
    Number of synthetic points added per instance mask. The 50/200 ratio is ablation-tested, but the total count is fixed by hand.
  • Gaussian standard deviations b1, b2 = not specified
    Appear in Eq. 3 but are never given in the implementation details, so the exact sampling distribution is unreported.
assumptions (3)
  • domain assumption The instance masks predicted by the segmentation network accurately delineate objects of interest.
    The whole RHGM samples points only inside these masks; erroneous masks directly contaminate the generated point cloud. Invoked in Section 'Point Cloud Generation' under 'Obtaining foreground points'.
  • domain assumption The depth and physical features of the nearest foreground point are a valid proxy for each generated point.
    In 'Generating hybrid points', every generated point inherits depth and features from the closest foreground point, which presumes that neighbor foreground point lies on the same object.
  • domain assumption Radar DOA estimation errors make true points more likely to lie near the estimated foreground points, with a Gaussian distribution governed by b1, b2.
    Eq. 3 models the generation PDF as Gaussian around foreground points; this is a heuristic with no quantitative justification or fitted error model.

how reviews work

0 comments
Cite this review

Pith. "Pith review of HGSFusion: Radar-Camera Fusion with Hybrid Generation and Synchronization for 3D Object Detection." pith.science (2026). https://pith.science/paper/Y654EP66

@misc{pith2026241211489,
  author       = {Pith},
  title        = {Pith review of: HGSFusion: Radar-Camera Fusion with Hybrid Generation and Synchronization for 3D Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Y654EP66}},
  note         = {Machine review of arXiv:2412.11489}
}
abstract

Millimeter-wave radar plays a vital role in 3D object detection for autonomous driving due to its all-weather and all-lighting-condition capabilities for perception. However, radar point clouds suffer from pronounced sparsity and unavoidable angle estimation errors. To address these limitations, incorporating a camera may partially help mitigate the shortcomings. Nevertheless, the direct fusion of radar and camera data can lead to negative or even opposite effects due to the lack of depth information in images and low-quality image features under adverse lighting conditions. Hence, in this paper, we present the radar-camera fusion network with Hybrid Generation and Synchronization (HGSFusion), designed to better fuse radar potentials and image features for 3D object detection. Specifically, we propose the Radar Hybrid Generation Module (RHGM), which fully considers the Direction-Of-Arrival (DOA) estimation errors in radar signal processing. This module generates denser radar points through different Probability Density Functions (PDFs) with the assistance of semantic information. Meanwhile, we introduce the Dual Sync Module (DSM), comprising spatial sync and modality sync, to enhance image features with radar positional information and facilitate the fusion of distinct characteristics in different modalities. Extensive experiments demonstrate the effectiveness of our approach, outperforming the state-of-the-art methods in the VoD and TJ4DRadSet datasets by $6.53\%$ and $2.03\%$ in RoI AP and BEV AP, respectively. The code is available at https://github.com/garfield-cpp/HGSFusion.

Figures

Figures reproduced from arXiv: 2412.11489 by the authors.

Figure 1
Figure 1. Illustration of angle estimation errors in obtaining [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overall framework of the proposed HGSFusion. In the radar branch, the RHGM utilizes raw radar points and images to [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Different encoding strategies of RHGM. Gener [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (5 more)
Figure 3
Figure 3. Figure 3: Point cloud generation in RHGM. Initially, raw [PITH_FULL_IMAGE:figures/full_fig_p004_3.png]
Figure 5
Figure 5. Figure 5: Internal structure of DSM. In Spatial Sync, radar features enhance image features with position information in radar [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Performance of different generation scheme. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Sample of the VoD dataset. “Unused Bicyle” is [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 8
Figure 8. Figure 8: Visualization results on the test set of TJ4DRadSet dataset under various lightning conditions. “Dark”, “Normal”, and [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

54 extracted references · 32 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Y.; and Ball, J

    Alaba, S. Y.; and Ball, J. E. 2022. A survey on deep-learning-based lidar 3d object detection for autonomous driving. Sensors, 22(24): 9577

    work page 2022

  4. [4]

    Aung, N. H. H.; Sangwongngam, P.; Jintamethasawat, R.; Shah, S.; and Wuttisittikulkij, L. 2024. A Review of LiDAR-based 3D Object Detection via Deep Learning Approaches towards Robust Connected and Autonomous Vehicles. IEEE Transactions on Intelligent Vehicles

    work page 2024

  5. [5]

    S.; Vegamoor, V

    Bhadoriya, A. S.; Vegamoor, V. K.; and Rathinam, S. 2021. Object detection and tracking for autonomous vehicles in adverse weather conditions. Technical report, SAE Technical Paper

    work page 2021

  6. [6]

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

    2023

  7. [7]

    G.; Kirillov, A.; and Girdhar, R

    Cheng, B.; Misra, I.; Schwing, A. G.; Kirillov, A.; and Girdhar, R. 2022. Masked-attention mask transformer for universal image segmentation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 1290--1299

    2022

  8. [8]

    Dong, X.; Wang, P.; Zhang, P.; and Liu, L. 2020. Probabilistic oriented object detection in automotive radar. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 102--103

    work page 2020

Show all 54 references
  1. [9]

    Dreher, M.; Er c elik, E.; B \"a nziger, T.; and Knoll, A. 2020. Radar-based 2D car detection using deep neural networks. In 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC), 1--8. IEEE

    2020

  2. [10]

    Ghasemieh, A.; and Kashef, R. 2022. 3D object detection for autonomous driving: Methods, models, sensors, data, and challenges. Transportation Engineering, 8: 100115

    2022

  3. [11]

    W.; Fang, Z.; Li, J.; Ambrus, R.; and Fragkiadaki, K

    Harley, A. W.; Fang, Z.; Li, J.; Ambrus, R.; and Fragkiadaki, K. 2023. Simple-bev: What really matters for multi-sensor bev perception? In 2023 IEEE International Conference on Robotics and Automation (ICRA), 2759--2765. IEEE

    2023

  4. [12]

    Hu, H.; Wang, F.; Su, J.; Wang, Y.; Hu, L.; Fang, W.; Xu, J.; and Zhang, Z. 2023. EA-LSS: Edge-aware Lift-splat-shot Framework for 3D BEV Object Detection. arXiv preprint arXiv:2303.17895, 2

    2023 arXiv

  5. [13]

    S.; Kuai, T.; and Waslander, S

    Hu, J. S.; Kuai, T.; and Waslander, S. L. 2022. Point density-aware voxels for lidar 3d object detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 8469--8478

    2022

  6. [14]

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

    2021 arXiv

  7. [15]

    Huang, K.-C.; Lyu, W.; Yang, M.-H.; and Tsai, Y.-H. 2024. PTT: Point-Trajectory Transformer for Efficient Temporal 3D Object Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 14938--14947

    2024

  8. [16]

    Jiao, Y.; Jie, Z.; Chen, S.; Chen, J.; Ma, L.; and Jiang, Y.-G. 2023. Msmdfusion: Fusing lidar and camera at multiple scales with multi-depth seeds for 3d object detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 21643--21652

    2023

  9. [17]

    W.; and Kum, D

    Kim, Y.; Kim, S.; Choi, J. W.; and Kum, D. 2023 a . Craft: Camera-radar 3d object detection with spatio-contextual fusion transformer. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, 1160--1168

    2023

  10. [18]

    W.; and Kum, D

    Kim, Y.; Shin, J.; Kim, S.; Lee, I.-J.; Choi, J. W.; and Kum, D. 2023 b . Crn: Camera radar net for accurate, robust, efficient 3d perception. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 17615--17626

    2023

  11. [19]

    K \"o hler, D.; Quach, M.; Ulrich, M.; Meinl, F.; Bischoff, B.; and Blume, H. 2023. Improved multi-scale grid rendering of point clouds for radar object detection networks. In 2023 26th International Conference on Information Fusion (FUSION), 1--8. IEEE

    2023

  12. [20]

    H.; Vora, S.; Caesar, H.; Zhou, L.; Yang, J.; and Beijbom, O

    Lang, A. H.; Vora, S.; Caesar, H.; Zhou, L.; Yang, J.; and Beijbom, O. 2019. Pointpillars: Fast encoders for object detection from point clouds. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 12697--12705

    2019

  13. [21]

    Li, H.; and Qu, H. 2024. DASSF: Dynamic-Attention Scale-Sequence Fusion for Aerial Object Detection. arXiv preprint arXiv:2406.12285

    2024 arXiv

  14. [22]

    Li, J.; Luo, C.; and Yang, X. 2023. PillarNeXt: Rethinking network designs for 3D object detection in LiDAR point clouds. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 17567--17576

    2023

  15. [23]

    Li, Y.; Ge, Z.; Yu, G.; Yang, J.; Wang, Z.; Shi, Y.; Sun, J.; and Li, Z. 2023. Bevdepth: Acquisition of reliable depth for multi-view 3d object detection. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, 1477--1485

    2023

  16. [24]

    Li, Z.; Wang, F.; and Wang, N. 2021. Lidar r-cnn: An efficient and universal 3d object detector. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 7546--7555

    2021

  17. [25]

    Lin, Z.; Liu, Z.; Xia, Z.; Wang, X.; Wang, Y.; Qi, S.; Dong, Y.; Dong, N.; Zhang, L.; and Zhu, C. 2024. 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, 14928--14937

    2024

  18. [26]

    Liu, J.; Zhao, Q.; Xiong, W.; Huang, T.; Han, Q.-L.; and Zhu, B. 2024 a . SMURF: Spatial multi-representation fusion for 3D object detection with 4D imaging radar. IEEE Transactions on Intelligent Vehicles

    2024

  19. [27]

    Liu, Y.; Wang, F.; Wang, N.; and ZHANG, Z.-X. 2024 b . Echoes beyond points: Unleashing the power of raw radar data in multi-modality fusion. Advances in Neural Information Processing Systems, 36

    2024

  20. [28]

    L.; and Han, S

    Liu, Z.; Tang, H.; Amini, A.; Yang, X.; Mao, H.; Rus, D. L.; and Han, S. 2023. Bevfusion: Multi-task multi-sensor fusion with unified bird's-eye view representation. In 2023 IEEE international conference on robotics and automation (ICRA), 2774--2781. IEEE

    2023

  21. [29]

    Ma, X.; Ouyang, W.; Simonelli, A.; and Ricci, E. 2023. 3d object detection from images for autonomous driving: a survey. IEEE Transactions on Pattern Analysis and Machine Intelligence

    2023

  22. [30]

    Mao, J.; Shi, S.; Wang, X.; and Li, H. 2023. 3D object detection for autonomous driving: A comprehensive survey. International Journal of Computer Vision, 131(8): 1909--1963

    2023

  23. [31]

    Meng, Z.; Xia, X.; Xu, R.; Liu, W.; and Ma, J. 2023. HYDRO-3D: Hybrid object detection and tracking for cooperative perception using 3D LiDAR. IEEE Transactions on Intelligent Vehicles, 8(8): 4069--4080

    2023

  24. [32]

    Paek, D.-H.; Kong, S.-H.; and Wijaya, K. T. 2022. K-radar: 4d radar object detection for autonomous driving in various weather conditions. Advances in Neural Information Processing Systems, 35: 3819--3829

    2022

  25. [33]

    Palffy, A.; Pool, E.; Baratam, S.; Kooij, J. F. P.; and Gavrila, D. M. 2022. Multi-Class Road User Detection With 3+1D Radar in the View-of-Delft Dataset. IEEE Robotics and Automation Letters, 7(2): 4961--4968

    2022

  26. [34]

    Philion, J.; and Fidler, S. 2020. 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, 194--210. Springer

    2020

  27. [35]

    Popov, A.; Gebhardt, P.; Chen, K.; and Oldja, R. 2023. Nvradarnet: Real-time radar obstacle and free space detection for autonomous driving. In 2023 IEEE International Conference on Robotics and Automation (ICRA), 6958--6964. IEEE

    2023

  28. [36]

    Reading, C.; Harakeh, A.; Chae, J.; and Waslander, S. L. 2021. Categorical depth distribution network for monocular 3d object detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 8555--8564

    2021

  29. [37]

    Rebut, J.; Ouaknine, A.; Malik, W.; and P \'e rez, P. 2022. Raw high-definition radar for multi-task learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 17021--17030

    2022

  30. [38]

    Shi, G.; Li, R.; and Ma, C. 2022. Pillarnet: Real-time and high-performance pillar-based 3d object detection. In European Conference on Computer Vision, 35--52. Springer

    2022

  31. [39]

    St \"a cker, L.; Heidenreich, P.; Rambach, J.; and Stricker, D. 2022. Fusion point pruning for optimized 2d object detection with radar-camera fusion. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 3087--3094

    2022

  32. [40]

    Tan, B.; Ma, Z.; Zhu, X.; Li, S.; Zheng, L.; Chen, S.; Huang, L.; and Bai, J. 2022 a . 3-D object detection for multiframe 4-D automotive millimeter-wave radar point cloud. IEEE Sensors Journal, 23(11): 11125--11138

    2022

  33. [41]

    Tan, B.; Ma, Z.; Zhu, X.; Li, S.; Zheng, L.; Chen, S.; Huang, L.; and Bai, J. 2022 b . 3d object detection for multi-frame 4d automotive millimeter-wave radar point cloud. IEEE Sensors Journal

    2022

  34. [42]

    o hler, D.; Niederl \

    Ulrich, M.; Braun, S.; K \"o hler, D.; Niederl \"o hner, D.; Faion, F.; Gl \"a ser, C.; and Blume, H. 2022. Improved orientation estimation and detection with hybrid object detection networks for automotive radar. In 2022 IEEE 25th International Conference on Intelligent Trans...

    2022

  35. [43]

    Wu, D.; Yang, F.; Xu, B.; Liao, P.; and Liu, B. 2024. A Survey of Deep Learning Based Radar and Vision Fusion for 3D Object Detection in Autonomous Driving. arXiv preprint arXiv:2406.00714

    2024 arXiv

  36. [44]

    Xiong, W.; Liu, J.; Huang, T.; Han, Q.-L.; Xia, Y.; and Zhu, B. 2023. LXL: LiDAR excluded lean 3D object detection with 4D imaging radar and camera fusion. IEEE Transactions on Intelligent Vehicles

    2023

  37. [45]

    Xu, B.; Zhang, X.; Wang, L.; Hu, X.; Li, Z.; Pan, S.; Li, J.; and Deng, Y. 2021. RPFA-Net: A 4D radar pillar feature attention network for 3D object detection. In 2021 IEEE International Intelligent Transportation Systems Conference (ITSC), 3061--3066. IEEE

    2021

  38. [46]

    Yan, J.; Liu, Y.; Sun, J.; Jia, F.; Li, S.; Wang, T.; and Zhang, X. 2023. Cross modal transformer: Towards fast and robust 3d object detection. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 18268--18278

    2023

  39. [47]

    Yan, Q.; and Wang, Y. 2023. Mvfan: Multi-view feature assisted network for 4d radar object detection. In International Conference on Neural Information Processing, 493--511. Springer

    2023

  40. [48]

    Yang, Z.; Chen, J.; Miao, Z.; Li, W.; Zhu, X.; and Zhang, L. 2022. Deepinteraction: 3d object detection via modality interaction. Advances in Neural Information Processing Systems, 35: 1992--2005

    2022

  41. [49]

    a henb \

    Yin, T.; Zhou, X.; and Kr \"a henb \"u hl, P. 2021. Multimodal virtual point 3d detection. Advances in Neural Information Processing Systems, 34: 16494--16507

    2021

  42. [50]

    Zhang, H.; Liang, L.; Zeng, P.; Song, X.; and Wang, Z. 2024 a . SparseLIF: High-Performance Sparse LiDAR-Camera Fusion for 3D Object Detection. arXiv preprint arXiv:2403.07284

    2024 arXiv

  43. [51]

    Zhang, H.; Wu, K.; Chen, R.; Wu, Z.; Zhong, Y.; and Li, W. 2024 b . TL-4DRCF: A two-level 4D radar-camera fusion method for object detection in adverse weather. IEEE Sensors Journal

    2024

  44. [52]

    Zhang, Y.; Hu, Q.; Xu, G.; Ma, Y.; Wan, J.; and Guo, Y. 2022. Not all points are equal: Learning highly efficient point-based detectors for 3d lidar point clouds. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 18953--18962

    2022

  45. [53]

    Zheng, L.; Li, S.; Tan, B.; Yang, L.; Chen, S.; Huang, L.; Bai, J.; Zhu, X.; and Ma, Z. 2023. Rcfusion: Fusing 4d radar and camera with bird’s-eye view features for 3d object detection. IEEE Transactions on Instrumentation and Measurement

    2023

  46. [54]

    Zheng, L.; Ma, Z.; Zhu, X.; Tan, B.; Li, S.; Long, K.; Sun, W.; Chen, S.; Zhang, L.; Wan, M.; Huang, L.; and Bai, J. 2022. TJ4DRadSet: A 4D Radar Dataset for Autonomous Driving. In 2022 IEEE 25th International Conference on Intelligent Transportation Systems (ITSC), 493--498

    2022

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.