Pith. sign in

REVIEW 4 major objections 5 minor 60 references

Radar Tracker: Moving Instance Tracking in Sparse and Noisy Radar Point Clouds

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A tracker that fuses per-point temporal offsets with attention-based appearance features achieves state-of-the-art moving-instance tracking in sparse radar point clouds.

desk verdict Solid radar tracking method, but the SOTA claim rests on an excluded baseline and single-run numbers. read the letter →

arxiv 2507.03441 v1 pith:5PK3YQ2O submitted 2025-07-04 cs.CV

Matthias Zeller , Daniel Casado Herraez , Jens Behley , Michael Heidingsfeld , Cyrill Stachniss This is my paper · ORCID
classification cs.CV
keywords movinginstancetrackingradarpointcloudstemporaloffsetpredictionattention-basedassociationcenter-basedScenes4Dpanopticsegmentationdata
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

This paper tries to establish that moving object tracking in sparse, noisy radar point clouds can be made reliable by combining two kinds of evidence: geometric motion cues and learned appearance features. It proposes a tracker that predicts, for every radar point, a temporal offset pointing to where its instance will be in the next scan, enabling direct center-based association even when an instance is a single point. On top of that, it adds an attention-based module that extracts appearance features and computes similarity scores between tracks and new detections. On the RadarScenes benchmark, this combination reaches an LSTQ of 66.8, about 7.5 points above a center-tracking baseline, which matters because radar remains the sensor that works in bad weather and its data is exceptionally sparse.

What carries the argument

The central mechanism is the per-point temporal offset prediction: for each radar point, the network regresses a vector to the center of its instance in the next scan, computed as the mean of the ground-truth points of that instance. Together with the standard offset to the current-scan center, this creates direct center-based tracks without voxelization or scan aggregation. The second mechanism is the attentive instance network, which applies local vector attention (using the Point Transformer's subtractive attention with relative positional encoding) to the moving points of each instance and aggregates them into a single instance feature vector. A similarity module then computes an attention matrix $A_{\text{sim}} = \text{sigmoid}(Q_{\text{sim}}K_{\text{sim}}^\top + R_{\text{sim}})$ that is converted into an appearance cost, and the data association clusters instances with DBSCAN and runs Hungarian matching locally, using the appearance cost only beyond a short-distance threshold.

What would settle it

Run the same tracker on radar sequences where most moving instances are single-point detections and compare the association score against a version that replaces the temporal-offset target with a velocity-based constant-motion prediction. If the association score drops to the level of the geometric-only baseline, the noisy next-scan center target is the bottleneck; alternatively, measuring track ID switches on scenes with two closely spaced pedestrians would test whether the appearance cost actually resolves geometric ambiguity.

Watch

Extended reading notes

Core claim

The paper's central claim is that a tracking-by-detection system for radar point clouds should associate instances using a cascade of geometric and appearance information rather than either alone. The geometric stage uses per-point temporal offsets to predict the next-scan center of each moving instance, making center-based Hungarian association viable in data where instances often consist of one or a few points. The appearance stage uses a transformer-based instance feature extractor and a sigmoid attention score to measure how likely two instances are the same object, which resolves the cases where geometry is ambiguous or objects interact. The final association switches from geometry-only to geometry-plus-appearance above a distance threshold, and the paper reports that this combined design outperforms the compared baselines on the RadarScenes moving-instance tracking benchmark.

Load-bearing premise

The method assumes that the average coordinate of the ground-truth points of an instance in the next scan is a reliable center target for every moving point, including instances that appear as a single point in the current scan.

Editorial extensions

If this is right

  • The temporal offset prediction adds motion cues that improve moving-object segmentation IoU from 84.4 to 85.4 on the validation set, and the center-based association alone gains 0.7 points in association score.
  • Attention-based appearance association is what closes long-distance or ambiguous matches: removing it and using only geometry at the 10 m threshold drops the association score from 54.3 to 52.2.
  • The method operates on single scans rather than aggregated scans, so it avoids the latency of scan accumulation, which matters for safety-critical tasks like collision avoidance.
  • Because the backbone is a separate moving-instance segmenter, the tracking modules can be attached to any other segmentation network, not only the one used here.
  • On the RadarScenes test set, the approach reaches LSTQ 66.8 with an association score of 48.2, versus 59.3 for center tracking plus Hungarian, indicating the gain is concentrated in association quality rather than classification.

Reading between the lines

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

  • If the combined association generalizes, radar-only perception stacks could provide reliable tracking in weather that disables cameras and LiDAR, since the method needs no dense point clouds or scan accumulation.
  • The temporal-offset target (mean of next-scan instance points) is a noisy label for single-point instances; a testable extension would be to weight the offset loss by instance point count or predict a distribution over next centers instead of a single vector.
  • The appearance similarity module could be trained with contrastive or metric losses to make instance features more discriminative, potentially pushing the association score further without changing the association logic.
  • The reported numbers are on a single benchmark with one sensor setup; a natural falsifier is running the same pipeline on another radar dataset to see whether the 7.5-point gap over center-only tracking persists across sensor configurations.
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 paper addresses moving instance tracking in sparse and noisy automotive radar point clouds. It extends the authors' Radar Instance Transformer (RIT) with two tracking-specific modules: a per-point temporal offset head that regresses each moving point to the center of its instance in the next scan, and an attention-based instance feature extractor that produces an appearance similarity cost. The final data association clusters instances with DBSCAN, performs Hungarian matching within clusters, and switches between geometric-only and combined geometric-appearance costs according to hand-set distance thresholds. Experiments on RadarScenes report LSTQ 66.8 with S_assoc 48.2, outperforming three implemented baselines, and ablations on the validation set support the contributions of the offset predictions and the attentive association.

Significance. If the experimental claims hold, the paper provides a sensible point-based alternative to voxelized or IoU-based trackers for radar, with a plausible mechanism for single-point instances (temporal offsets) and a modest but consistent improvement in association quality. The paper's strengths include evaluation on a public benchmark with a standard metric, implementation details sufficient for reproduction, and a set of ablations that isolate the main design choices. The main caveats are that the comparative claim rests on a selected baseline set, all numbers come from a single run, and several association thresholds are set manually without sensitivity analysis; these issues do not invalidate the method but currently temper the strength of the 'state-of-the-art' statement.

major comments (4)
  1. [Sec. IV-B, Table I] The abstract and Sec. I state that the approach achieves 'state-of-the-art performance' for moving instance tracking, but Table I compares only MOT [46], center tracking [49], and CA-Net [26]. The best-performing 4D panoptic method, Eq-4D-StOP [59], is excluded with the statement that its rotation augmentation is 'detrimental to radar data [29]'. The cited reference (View-of-Delft, Palffy et al.) does not evaluate Eq-4D-StOP, and no experiment quantifies the effect of rotation augmentation on this method. Because the headline claim is comparative, this exclusion is load-bearing. Please either add a fair adaptation of Eq-4D-StOP to the RadarScenes evaluation, provide direct evidence that it cannot be transferred, or rephrase the claim to 'improved performance over the implemented baselines'.
  2. [Sec. IV-A, Tables I-III] All results are reported for a single run, with no variance or significance measures. The ablation differences in Tables II and III are small (0.1-0.3 LSTQ or S_assoc points), which is within the range of typical seed-to-seed variation for transformer-based point cloud models. The main 7.5 LSTQ gap over center tracking (Table I) is therefore not yet distinguishable from evaluation-protocol or initialization effects. Please report mean and standard deviation over at least three seeds, or provide an alternative statistical analysis.
  3. [Sec. III-F] The data association contains several hand-set free parameters: the DBSCAN bandwidth b=10, the geometric/appearance thresholds t_d1=5 m and t_d2=10 m, the similarity cost threshold t_c=1.5, and the track retention length of 12 scans. No sensitivity analysis is provided. These parameters directly determine when the appearance cost is used and when tracks are initialized or dropped, and their values may have been chosen with knowledge of the validation set. Please include a sensitivity study over the main thresholds (at least t_d1, t_d2, t_c, b) and state explicitly whether any parameter was tuned on the test set.
  4. [Sec. III-B] The temporal offset target is defined as the mean coordinate of the ground-truth instance points in the next scan. For instances consisting of a single radar detection, this target is a single noisy point whose position can jump arbitrarily between frames, and the current ablation (Table II) does not isolate this case. Since the paper motivates the center-based association partly by the single-point-instance scenario, please add an analysis or ablation that reports performance as a function of instance size (e.g., single-point vs. multi-point instances) or otherwise demonstrates stability of the temporal-offset target. Without such evidence, the claim in Sec. III-C that the temporal offset provides a regression target for single-point instances is not quantitatively supported.
minor comments (5)
  1. [Sec. III-B, Eq. (7)] The notation in Eq. (7) is ambiguous: o_i is used for both offsets, and c_i is said to be 'the respective center of the instance that p_i belongs to', but for the temporal offset this must be the center in the next scan. Please define the standard and temporal offsets with distinct symbols in the loss equation.
  2. [Fig. 2] The figure shows W_K and W_Q but not W_V in the attentive instance network, whereas Eq. (1) defines W_V; please align the figure with the text.
  3. [References] The same workshop paper 4D-StOP appears as two separate references [20] and [21]; please merge them into a single citation.
  4. [Sec. IV-B] The center tracking baseline description says Doppler velocities are 'directly used' for tracking, but the procedure is not described; please specify whether velocities are used as motion model predictions, as gating, or as association features.
  5. [Sec. IV-D, Table III] The row 'Geometric association t_d2=10 m' is unclear; please clarify what 'geometric association' means in this ablation, for example whether it is the same pipeline with the appearance cost removed or a pure distance threshold without the attentive similarity.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the central tracking claim is an external benchmark comparison on the RadarScenes test set, and the learned offset and similarity components are supervised on ground-truth targets rather than fitted to the evaluation metric.

full rationale

The paper's main claim is comparative performance on the RadarScenes moving-instance tracking benchmark. The LSTQ metric is defined externally in 4D panoptic segmentation literature, and the reported test-set numbers are obtained by evaluating a trained model on held-out sequences. The temporal offset prediction in Sec. III-B is trained with the supervised L1 loss in Eq. (7) against ground-truth instance centers; the attention similarity in Eq. (5) is trained with a binary cross-entropy loss. Neither quantity is fitted to the LSTQ score or to the association threshold, so the 'prediction' is not equivalent to an input by construction. The paper does cite the authors' own Radar Instance Transformer [51] as the backbone and adopts its training settings, but RIT is prior published work with its own external evaluation on RadarScenes, and the baselines are instantiated with the same backbone to isolate the tracking contribution. This is a normal use of prior work rather than a load-bearing self-citation chain. The exclusion of Eq-4D-StOP [59] is a potential evidence gap for the 'state-of-the-art' wording, because the reason for exclusion is qualitative and cites a different radar dataset; however, that is a concern about comparison completeness and correctness risk, not circularity. Overall, no equation or fitted parameter reduces to the claimed result, and the central derivation is self-contained against an external benchmark.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The reported performance rests primarily on a learned network plus a set of hand-set association thresholds and dataset assumptions. The most consequential items are the temporal offset regression target (Sec. III-B), the DBSCAN and threshold parameters (Sec. III-F), and the decision to exclude Eq-4D-StOP from the benchmark comparison (Sec. IV-B).

free parameters (6)
  • similarity cost threshold t_c = 1.5
    Hand-set in Sec. III-F. Controls when the appearance-based cost is used in the association and when new tracks are initialized; no sensitivity analysis is provided.
  • distance threshold t_d1 = 5 m
    Hand-set in Sec. III-F. Below this distance the association is purely geometric, so it directly defines the operating regime of the geometric module.
  • distance threshold t_d2 = 10 m
    Hand-set in Sec. III-F. Above this distance the appearance-based association is considered too difficult and association is omitted.
  • DBSCAN bandwidth b = 10
    Hand-set in Sec. III-F. Determines local clustering of instances before Hungarian matching; affects which global false matches are pruned.
  • track retention length = 12 scans
    Hand-set in Sec. III-F. Determines how long occluded tracks are propagated; directly affects association and initialization of new tracks.
  • number of local attention neighbors N_l = 6
    Set in Sec. III-F for the k-nearest-neighbor grouping in the Point Transformer blocks; trades computational cost and local context.
assumptions (6)
  • standard math Hungarian algorithm returns an optimal bipartite matching for the local cost matrices.
    Invoked in Sec. III-E as the association solver; correctness is standard.
  • domain assumption DBSCAN clustering correctly groups nearby instances into local neighborhoods that contain no misleading global matches.
    Assumed in Sec. III-E; the bandwidth b=10 is fixed without sensitivity analysis and could split or merge clusters in sparse radar scenes.
  • domain assumption RadarScenes ground-truth instance IDs are temporally consistent, and the mean of instance points in the next scan is a valid regression target for the temporal offset.
    Used in Sec. III-B to define the temporal offset loss in Eq. (7); if annotations are inconsistent or centers are unstable, the offset targets are noisy.
  • ad hoc to paper Excluding Eq-4D-StOP from the comparison because it uses large rotations of input point clouds is a valid reason, and no adaptation of that method to radar would change the ranking.
    Stated in Sec. IV-B; this assumption is load-bearing for the state-of-the-art claim but is not tested.
  • domain assumption Training the model with ground-truth instance selection and evaluating it with predicted instances does not create a harmful train/evaluation mismatch.
    Described in Sec. III-C; the model sees clean instance masks during training but noisy predicted masks at inference, which can degrade association performance.
  • domain assumption LSTQ score adapted from [2] is a valid metric for radar moving-instance tracking, and scores across methods using the same backbone are comparable.
    Stated in Sec. IV-A; comparability depends on identical evaluation scripts and backbone initialization.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Radar Tracker: Moving Instance Tracking in Sparse and Noisy Radar Point Clouds." pith.science (2026). https://pith.science/paper/5PK3YQ2O

@misc{pith2026250703441,
  author       = {Pith},
  title        = {Pith review of: Radar Tracker: Moving Instance Tracking in Sparse and Noisy Radar Point Clouds},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5PK3YQ2O}},
  note         = {Machine review of arXiv:2507.03441}
}
read the original abstract

Robots and autonomous vehicles should be aware of what happens in their surroundings. The segmentation and tracking of moving objects are essential for reliable path planning, including collision avoidance. We investigate this estimation task for vehicles using radar sensing. We address moving instance tracking in sparse radar point clouds to enhance scene interpretation. We propose a learning-based radar tracker incorporating temporal offset predictions to enable direct center-based association and enhance segmentation performance by including additional motion cues. We implement attention-based tracking for sparse radar scans to include appearance features and enhance performance. The final association combines geometric and appearance features to overcome the limitations of center-based tracking to associate instances reliably. Our approach shows an improved performance on the moving instance tracking benchmark of the RadarScenes dataset compared to the current state of the art.

Figures

Figures reproduced from arXiv: 2507.03441 by the authors.

Figure 1
Figure 1. Our method combines moving object segmentation (top), [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The detailed design of the individual modules of our Radar Tracker. (a) The backbone is extended with the offset predictions and [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 43 canonical work pages

  1. [29]

    Palffy, E

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

    work page 2022

  2. [46]

    X. Weng, J. Wang, D. Held, and K. Kitani. 3d multi-object tracking: A baseline and new evaluation metrics. In Proc. of the IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS) , 2020

    work page 2020

  3. [49]

    T. Yin, X. Zhou, and P. Kr ¨ahenb¨uhl. Center-Based 3D Object Detection and Tracking. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) , 2021

    work page 2021

  4. [26]

    Marcuzzi, L

    R. Marcuzzi, L. Nunes, L. Wiesmann, I. Vizzo, J. Behley, and C. Stachniss. Contrastive Instance Association for 4D Panoptic Segmentation for Sequences of 3D LiDAR Scans. In Proc. of the IEEE Intl. Conf. on Robotics & Automation (ICRA) , 2022

    work page 2022

  5. [59]

    M. Zhu, S. Han, H. Cai, S. Borse, M.G. Jadidi, and F. Porikli. 4D Panoptic Segmentation as Invariant and Equivariant Field Prediction. arXiv preprint, arXiv:2303.15651, 2023

    work page Pith review arXiv 2023

  6. [1]

    Agarwalla, X

    A. Agarwalla, X. Huang, J. Ziglar, F. Ferroni, L. Leal-Taix ´e, J. Hays, A. Osep, and D. Ramanan. Lidar Panoptic Segmentation and Tracking without Bells and Whistles. In Proc. of the IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS) , 2023

    work page 2023

  7. [2]

    Ayg ¨un, A

    M. Ayg ¨un, A. Osep, M. Weber, M. Maximov, C. Stachniss, J. Behley, and L. Leal-Taixe. 4D Panoptic Segmentation. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR), 2021

    work page 2021

  8. [3]

    Benbarka, J

    N. Benbarka, J. Schr ¨oder, and A. Zell. Score refinement for confidence-based 3D multi-object tracking. In Proc. of the IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS) , 2021

    work page 2021

Show all 60 references
  1. [4]

    X. Chen, S. Li, B. Mersch, L. Wiesmann, J. Gall, J. Behley, and C. Stachniss. Moving Object Segmentation in 3D LiDAR Data: A Learning-based Approach Exploiting Sequential Data. IEEE Robotics and Automation Letters (RA-L) , 6(4):6529–6536, 2021

    2021

  2. [5]

    Y . Chen, J. Liu, X. Zhang, X. Qi, and J. Jia. V oxelnext: Fully sparse voxelnet for 3d object detection and tracking. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR), 2023

    2023

  3. [6]

    H.K. Chiu, A. Prioletti, J. Li, and J. Bohg. Probabilistic 3d multi-object tracking for autonomous driving. arXiv preprint , arXiv:2001.05673, 2020

    2001 arXiv

  4. [7]

    C. Choy, J. Gwak, and S. Savarese. 4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) , 2019

    2019

  5. [8]

    An algorithm for tracking multiple targets

    D.B, Reid. An algorithm for tracking multiple targets. IEEE Trans. on Automatic Control , 24(6):843–854, 1979

    1979

  6. [9]

    Dubey, A

    A. Dubey, A. Santra, J. Fuchs, M. L ¨ubke, R. Weigel, and F. Lurz. HARadNet: Anchor-free target detection for radar point clouds using hierarchical attention and multi-task learning. Machine Learning with Applications (MLWA), 8:100275, 2022

    2022

  7. [10]

    Ester, H

    M. Ester, H. Kriegel, J. Sander, and X. Xu. A density-based algorithm for discovering clusters in large spatial databases with noise. In Proc. of the Conf. on Knowledge Discovery and Data Mining (KDD) , 1996

    1996

  8. [11]

    M.H. Guo, J. Cai, Z.N. Liu, T.J. Mu, R.R. Martin, and S. Hu. PCT: Point Cloud Transformer. Computational Visual Media, 7(2):187–199, 2021

    2021

  9. [12]

    F. Hong, H. Zhou, X. Zhu, H. Li, and Z. Liu. Lidar-based panoptic segmentation via dynamic shifting network. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) , 2021

    2021

  10. [13]

    F. Hong, H. Zhou, X. Zhu, H. Li, and Z. Liu. Lidar-based 4d panoptic segmentation via dynamic shifting network. arXiv preprint , arXiv:2203.07186, 2022

    2022 arXiv

  11. [14]

    Huang, Z

    S. Huang, Z. Gojcic, J. Huang, A. Wieser, and K. Schindler. Dynamic 3D Scene Analysis by Point Cloud Accumulation. In Proc. of the Europ. Conf. on Computer Vision (ECCV) , 2022

    2022

  12. [15]

    L. Hui, L. Wang, L. Tang, K. Lan, J. Xie, and J. Yang. 3D Siamese transformer network for single object tracking on point clouds. In Proc. of the Europ. Conf. on Computer Vision (ECCV) , 2022

    2022

  13. [16]

    Ioffe and C

    S. Ioffe and C. Szegedy. Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift. In Proc. of the Intl. Conf. on Machine Learning (ICML) , 2015

    2015

  14. [17]

    R. Kalman. A New Approach to Linear Filtering and Prediction Problems. Transactions of the ASME – Journal of Basic Engineering , 82:35–45, 1960

    1960

  15. [18]

    D. Kim, S. Woo, J. Lee, and I.S. Kweon. Video Panoptic Segmenta- tion. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) , 2020

    2020

  16. [19]

    J. Kim, J. Woo, and Sunghoon. RVMOS: Range-View Moving Object Segmentation Leveraged by Semantic and Motion Features. IEEE Robotics and Automation Letters (RA-L) , 7(3):8044–8051, 2022

    2022

  17. [20]

    Kreuzberg, I.E

    L. Kreuzberg, I.E. Zulfikar, S. Mahadevan, F. Engelmann, and B. Leibe. 4d-stop: Panoptic segmentation of 4d lidar using spatio- temporal object proposal generation and aggregation. In Proc. of the Europ. Conf. on Computer Vision (ECCV) , 2022

    2022

  18. [21]

    Kreuzberg, I.E

    L. Kreuzberg, I.E. Zulfikar, S. Mahadevan, F. Engelmann, and B. Leibe. 4D-StOP: Panoptic Segmentation of 4D LiDAR using Spatio-temporal Object Proposal Generation and Aggregation. In Proc. of the Europ. Conf. on Computer Vision Workshops , 2022

    2022

  19. [22]

    H. Kuhn. The hungarian method for the assignment problem. Naval Research Logistics Quarterly , 2(1-2):83–97, 1955

    1955

  20. [23]

    J. Li, X. He, Y . Wen, Y . Gao, X. Cheng, and D. Zhang. Panoptic-phnet: Towards real-time and high-precision lidar panoptic segmentation via clustering pseudo heatmap. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) , 2022

    2022

  21. [24]

    Loshchilov and F

    I. Loshchilov and F. Hutter. SGDR: Stochastic Gradient Descent with Warm Restarts. In Proc. of the Intl. Conf. on Learning Representations (ICLR), 2017

    2017

  22. [25]

    Marcuzzi, L

    R. Marcuzzi, L. Nunes, L. Wiesmann, J. Behley, and C. Stachniss. Mask-Based Panoptic LiDAR Segmentation for Autonomous Driving. IEEE Robotics and Automation Letters (RA-L) , 8(2):1141–1148, 2023

    2023

  23. [27]

    Mersch, X

    B. Mersch, X. Chen, I. Vizzo, L. Nunes, J. Behley, and C. Stachniss. Receding Moving Object Segmentation in 3D LiDAR Data Using Sparse 4D Convolutions. IEEE Robotics and Automation Letters (RA- L), 7(3):7503–7510, 2022

    2022

  24. [28]

    Nair and G

    V . Nair and G. Hinton. Rectified linear units improve restricted boltzmann machines. In Proc. of the Intl. Conf. on Machine Learning (ICML), 2010

    2010

  25. [30]

    Z. Pang, Z. Li, and N. Wang. Simpletrack: Understanding and rethinking 3d multi-object tracking. In Proc. of the Europ. Conf. on Computer Vision (ECCV) , 2022

    2022

  26. [31]

    C. Park, Y . Jeong, M. Cho, and J. Park. Fast Point Transformer. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) , 2022

    2022

  27. [32]

    Paszke, S

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, et al. Pytorch: An imperative style, high-performance deep learning library. In Proc. of the Conf. on Neural Information Processing Systems (NeurIPS) , 2019

    2019

  28. [33]

    C.R. Qi, H. Su, K. Mo, and L.J. Guibas. PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation. In Proc. of the IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) , 2017

    2017

  29. [34]

    C. Qi, K. Yi, H. Su, and L.J. Guibas. PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space. In Proc. of the Conf. on Neural Information Processing Systems (NeurIPS) , 2017

    2017

  30. [35]

    H. Qiu, B. Yu, and D. Tao. GFNet: Geometric Flow Network for 3D Point Cloud Semantic Segmentation. Trans. on Machine Learning Research (TMLR), 2022

    2022

  31. [36]

    Schult, F

    J. Schult, F. Engelmann, A. Hermans, O. Litany, S. Tang, and B. Leibe. Mask3D: Mask Transformer for 3D Semantic Instance Segmentation. In Proc. of the IEEE Intl. Conf. on Robotics & Automation (ICRA) , 2023

    2023

  32. [37]

    Schumann, M

    O. Schumann, M. Hahn, J. Dickmann, and C. W ¨ohler. Semantic Segmentation on Radar Point Clouds. In Proc. of the Intl. Conf. on Information Fusion , 2018

    2018

  33. [38]

    Schumann, M

    O. Schumann, M. Hahn, N. Scheiner, F. Weishaupt, J.F. Tilly, J. Dick- mann, and C. W ¨ohler. RadarScenes: A real-world radar point cloud data set for automotive applications. In Proc. of the Intl. Conf. on Information Fusion , 2021

    2021

  34. [39]

    Schumann, J

    O. Schumann, J. Lombacher, M. Hahn, C. W ¨ohler, and J. Dickmann. Scene Understanding With Automotive Radar. IEEE Trans. on Intelligent V ehicles, 5(2):188–203, 2019

    2019

  35. [40]

    Shi and K

    Y . Shi and K. Ma. SAFIT: Segmentation-Aware Scene Flow with Improved Transformer. In Proc. of the IEEE Intl. Conf. on Robotics & Automation (ICRA) , 2022

    2022

  36. [41]

    S. Su, J. Xu, H. Wang, Z. Miao, X. Zhan, D. Hao, and X. Li. PUPS: Point cloud unified panoptic segmentation. arXiv preprint , arXiv:2302.06185, 2023

    2023 arXiv

  37. [42]

    J. Sun, Y . Dai, X. Zhang, J. Xu, R. Ai, W. Gu, and X. Chen. Efficient Spatial-Temporal Information Fusion for LiDAR-Based 3D Moving Object Segmentation. In Proc. of the IEEE/RSJ Intl. Conf. on Intelligent Robots and Systems (IROS) , 2022

    2022

  38. [43]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A.N. Gomez, Ł. Kaiser, and I. Polosukhin. Attention is all you need. In Proc. of the Conf. on Neural Information Processing Systems (NeurIPS), 2017

    2017

  39. [44]

    T. Vu, K. Kim, T.M. Luu, T. Nguyen, and C.D. Yoo. SoftGroup for 3D Instance Segmentation on Point Clouds. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) , 2022

    2022

  40. [45]

    N. Wang, C. Shi, R. Guo, H. Lu, Z. Zheng, and X. Chen. InsMOS: Instance-Aware Moving Object Segmentation in LiDAR Data. arXiv preprint, arXiv:2303.03909, 2023

    2023 arXiv

  41. [47]

    L. Xin, L. Jianhui, J. Li, W. Liwei, Z. Hengshuang, L. Shu, Q. Xi- aojuan, and J. Jiaya. Stratified Transformer for 3D Point Cloud Segmentation. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) , 2022

    2022

  42. [48]

    Xu, Y .C

    T.X. Xu, Y .C. Guo, Y .K. Lai, and S.H. Zhang. CXTrack: Improving 3D point cloud tracking with contextual information. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR), 2023

    2023

  43. [50]

    Zeller, J

    M. Zeller, J. Behley, M. Heidingsfeld, and C. Stachniss. Gaussian Radar Transformer for Semantic Segmentation in Noisy Radar Data. IEEE Robotics and Automation Letters (RA-L) , 8(1):344–351, 2023

    2023

  44. [51]

    Zeller, V .S

    M. Zeller, V .S. Sandhu, B. Mersch, J. Behley, M. Heidingsfeld, and C. Stachniss. Radar Instance Transformer: Reliable Moving Instance Segmentation in Sparse Radar Point Clouds. IEEE Trans. on Robotics (TRO), pages 1–17, 2023

    2023

  45. [52]

    Zeller, V .S

    M. Zeller, V .S. Sandhu, B. Mersch, J. Behley, M. Heidingsfeld, and C. Stachniss. Radar Velocity Transformer: Single-scan Moving Object Segmentation in Noisy Radar Point Clouds. In Proc. of the IEEE Intl. Conf. on Robotics & Automation (ICRA) , 2023

    2023

  46. [53]

    Zhang, C

    C. Zhang, C. Zhang, Y . Guo, L. Chen, and M. Happold. Motiontrack: End-to-end transformer-based multi-object tracking with lidar-camera fusion. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition Workshops, 2023

    2023

  47. [54]

    Zhang, H

    C. Zhang, H. Wan, X. Shen, and Z. Wu. PVT: Point-voxel trans- former for point cloud learning. Intl. Journal of Intelligent Systems , 37(12):11985–12008, 2022

    2022

  48. [55]

    Zhang, Z

    Y . Zhang, Z. Zhou, P. David, X. Yue, Z. Xi, B. Gong, and H. Foroosh. Polarnet: An improved grid representation for online lidar point clouds semantic segmentation. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) , 2020

    2020

  49. [56]

    H. Zhao, J. Jia, and V . Koltun. Exploring self-attention for image recognition. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) , 2020

    2020

  50. [57]

    H. Zhao, L. Jiang, J. Jia, P.H. Torr, and V . Koltun. Point Transformer. In Proc. of the IEEE/CVF Intl. Conf. on Computer Vision (ICCV) , 2021

    2021

  51. [58]

    Y . Zhou, L. Liu, H. Zhao, M. L ´opez-Ben´ıtez, L. Yu, and Y . Yue. Towards Deep Radar Perception for Autonomous Driving: Datasets, Methods, and Challenges. Sensors, 22, 2022

    2022

  52. [60]

    X. Zhu, H. Zhou, T. Wang, F. Hong, Y . Ma, W. Li, H. Li, and D. Lin. Cylindrical and asymmetrical 3d convolution networks for lidar segmentation. In Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR) , 2021

    2021

Pith tools

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