Fast2comm:Collaborative perception combined with prior knowledge

Hongkun Zhang; Yan Wu; Zhengbin Zhang

REVIEW 3 major objections 5 minor 26 references

Fast2comm:Collaborative perception combined with prior knowledge

T0 review · 3 major / 5 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 Fast2comm claims that collaborative perception improves when ground-truth bounding boxes supervise the confidence features that agents share, and that sharing fewer, target-gated features beats sharing top-k confidence features.

desk verdict GT-supervised feature selection is a reasonable idea, but the train/test decoupling is under-specified and the robustness claim needs better support. read the letter →

arxiv 2505.00740 v1 pith:DGEO337O submitted 2025-04-30 cs.CV cs.MA

Zhengbin Zhang , Yan Wu , Hongkun Zhang This is my paper · ORCID

classification cs.CVcs.MA

keywords collaborativeperceptionmulti-agentV2XautonomousdrivingBEVfeatureselectionconfidencemapbandwidthefficiencylocalizationrobustness3Dobjectdetection

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 proposes Fast2comm, a collaborative perception framework in which vehicles share a sparse set of Bird's Eye View features chosen by prior knowledge. It claims that supervising a confidence map with ground-truth labels, and using ground-truth bounding boxes to select which feature regions are shared, gives a better accuracy-bandwidth trade-off than the top-k confidence selection used by Where2comm. Across OPV2V, V2XSet, and DAIR-V2X, Fast2comm reports consistent gains over Where2comm, with the largest being 2.7 and 2.9 AP points on V2XSet, and it keeps an advantage when localization noise is added. If the claim holds, connected vehicles could transmit less data while detecting occluded or distant objects more reliably.

What carries the argument

The mechanism is a pair of feature-gating maps applied to the BEV feature map. The confidence map is produced by a detection head on an attention-fused feature representation, with an extra loss that supervises it against ground-truth labels so that it marks foreground. The second map is built by projecting each 7D ground-truth box into a 4D BEV box and binarizing it, so that it marks exactly the target regions. Both maps are applied by element-wise multiplication; agents transmit only non-zero entries, and the receiver concatenates the two gated tensors and fuses them with self-attention. At test time only the confidence-gated tensor is shared.

What would settle it

Train the same Fast2comm with the GT-box-gated features also shared at test time and measure AP; compare it with the paper's reported test-time setting that shares only confidence-gated features. If the accuracy drop from removing the GT-gated features is large, the reported gains depend on a training-only signal that real inference does not have.

Watch

Extended reading notes

Core claim

The central claim is that ground truth can be made to do double duty: it trains a confidence head to separate foreground from background, and it defines a binary BEV mask that restricts shared features to target regions. The paper shows that these two mechanisms are coupled—each alone hurts performance relative to the baseline, while together they improve it—and argues this is because the confidence map tells the mask where objects are, while the mask keeps background noise out of the fusion. It also claims that the decoupled fusion strategy (sharing both feature types during training, only the confidence-gated ones at test) preserves accuracy while cutting communication cost.

Load-bearing premise

The load-bearing premise is that the confidence map learned under ground-truth supervision selects the same target regions at test time that ground-truth boxes selected during training, and that the fusion module, trained on both feature types, still works when only the confidence-gated features arrive.

Editorial extensions

If this is right

Fast2comm reports gains over Where2comm of 1.0/1.2 AP points on OPV2V, 2.7/2.9 on V2XSet, and 1.5/0.9 on DAIR-V2X, with sparser messages.
Under simulated localization noise, the method's advantage over Where2comm grows (3.6 AP points at AP@0.5 with 0.5 m noise on OPV2V), indicating the prior-supervised confidence map helps absorb misalignment.
Because only non-zero selected features and their indices are transmitted, communication cost scales with sparsity; the decoupled training/testing design makes test-time messages smaller than training-time messages.
Ablations show the two proposed modules are synergistic: adding either alone lowers AP@0.7 on OPV2V and V2XSet, while adding both raises it above the baseline.

Reading between the lines

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

Editorial inference: the decoupled training/testing design suggests a cheap extension—at test time, select a connected-region mask from the confidence map instead of per-pixel top-k, which may recover part of the GT-gated signal without needing labels.
Editorial inference: the paper's robustness gain under GPS noise is consistent with the selection mask being anchored to object-sized regions rather than isolated high-confidence pixels; if so, any collaborative perception method that shares spatially coherent masks should show similar noise tolerance.
Editorial inference: the reported degradation when only one module is used implies that prior-supervised confidence generation and GT-box selection are coupled in the loss landscape; a practitioner adopting one without the other should expect a performance drop.

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.

Desk Editor's Note

read the letter

Colleague,

Read the Fast2comm paper. The short version: the core idea is neat, but the paper doesn't currently demonstrate that the reported gains survive the test-time setup.

What's actually new: they supervise the confidence map with ground-truth labels to make it sharper, and they use GT bounding boxes to select features during training only. At test, they only share the confidence-gated features M, not the GT-selected features G. That decoupling is the paper's main trick, and it's plausible—train with richer supervision, test with a cheaper message. The experiments span three datasets, the code is public, and the numbers beat Where2comm consistently.

The problem is the test-time behavior is under-specified. The fusion module (Eq. 5) is trained on concatenated M and G, but at test G is gone. The paper never says whether G is zero-padded, the sequence length changes, or the self-attention somehow copes with missing inputs. Since self-attention is length-agnostic but feature-dimension-sensitive, this matters a lot. If G is concatenated along channels, the model sees a different input dimension; if along spatial tokens, it sees half the context. Either way, the test-time input distribution differs from training. The authors might have handled this with a masking trick, but it's not in the paper.

The robustness claim is also partly tied to GT-FS, which isn't used at test. Section IV.C attributes the localization-robustness gains to both CFG and GT-FS, but GT-FS doesn't exist at inference. Those gains must come from the improved confidence map alone, and the ablation doesn't isolate that cleanly. In fact, Table II shows that adding GT-FS alone hurts a lot on OPV2V (78.44 to 71.82), which suggests the interaction is fragile.

The ablation is the most unusual part. CFG alone barely changes, GT-FS alone hurts, both together help. That's not intrinsically wrong—synergies happen—but the explanation in the paper is hand-wavy, and it doesn't address what happens at test when G is absent.

All that said, I don't think the paper is a wash. The prior-supervised confidence map is a sensible idea, and the reported improvements might be real, but the current write-up doesn't prove the test-time transfer. A good revision would specify the inference-time fusion, add an ablation where G is dropped at test, and retract the robustness attribution if it doesn't hold.

I'd send this to peer review rather than desk reject—it's a credible incremental contribution with code and multi-dataset evaluation—but the reviewers should focus on closing the train/test gap.

Referee Report

3 major / 5 minor

Summary. The paper proposes Fast2comm, a collaborative perception framework for multi-agent 3D object detection that combines prior-supervised confidence feature generation, GT bounding box-based feature selection, and a decoupled training/testing fusion strategy. The authors claim that Fast2comm achieves a better perception-communication trade-off and improved robustness to localization errors compared with prior methods, with reported gains over Where2comm of 1.0%/1.2% (OPV2V), 2.7%/2.9% (V2XSet), and 1.5%/0.9% (DAIR-V2X) in AP@0.5/0.7. The manuscript includes experiments on three datasets, ablation studies, communication-volume comparisons, and qualitative visualizations.

Significance. If the reported results hold at inference time, the idea of using ground-truth bounding boxes as a training-time structural prior to shape a sparse confidence-based communication channel is a useful contribution to collaborative perception, and the public code release is a positive step. The empirical study is broad, covering three datasets and including robustness and bandwidth analyses. However, the central claim is undermined by a train/test discrepancy: the fusion module is trained on concatenated confidence and GT-bbox features but at test time only confidence features are shared, and the paper does not specify how the trained fusion module handles the missing input. Because the reported gains are obtained in a configuration that is not the deployed one, the experimental validation as written does not establish the paper's main claims.

major comments (3)

[Section III.D, III.E, Eq. (5)] The paper states that during training both prior features G and confidence features M are shared, but during testing only M is shared. The fusion module in Eq. (5) applies self-attention to the flattened concatenation of M and G, so the trained model has never observed inputs containing only M during training. The manuscript does not specify how the self-attention input is re-formed at test time (zero-padding, masking, or a modified sequence length), nor does it report results for the test-time configuration. Since all quantitative comparisons in Tables I and Figures 5-6 are presumably obtained with a model that at test time receives only M, the reported accuracy and bandwidth figures cannot be taken at face value without a description and validation of this adaptation.
[Section IV.C.3, IV.E, Table II] The robustness improvement in Figure 6 is attributed to both the Confidence Feature Generation module and the GT Bbox-Based Feature Selection module, but GT-FS is a training-only mechanism: at test time the GT bbox-based branch is not shared. The ablation in Table II shows that GT-FS alone degrades OPV2V AP@0.7 from 78.44 to 71.82, and only the combination CFG+GT-FS exceeds the baseline. This combination is trained with both M and G, yet at test time G is absent; the manuscript does not demonstrate that the trained model retains this improvement when G is removed. The robustness claim therefore conflates training-time supervision with test-time feature selection, and the experiment needs to be rerun or explicitly reported in the actual test-time configuration.
[Section IV.E, Table II] The ablation study is underspecified regarding the training and testing protocol. It is unclear whether each ablation row uses the same training-time inputs (M and G) and only changes test-time inputs, or whether the presence of GT-FS changes the training procedure itself. In particular, the row with CFG only reports lower AP than the baseline, but if the baseline already used only M at test time, the comparison is not controlled for the fusion input. The authors should clarify how the train/test decoupling applies in each ablation row and provide results for the exact deployed setting (training with M and G, testing with M only) with appropriate architectural adaptation.

minor comments (5)

[Throughout] Typos and inconsistent naming appear in several places: 'F ast2comm' in the abstract, 'Faste2comm' in Section III.B.2, 'reprensents' in Eq. (2), 'Boudning' in the Section III.C heading, 'Robutness' in the Section IV.C.3 heading, 'hte' in Section III.G, and 'Where2Comm' with inconsistent capitalization. A careful proofread is needed.
[Section III.B.2] The confidence map threshold t in Eq. (1) is never defined or discussed in the experiments. The value of t directly controls the sparsity of M and hence the communication volume, so the authors should report the threshold used in all experiments and whether it is tuned per dataset.
[Section IV.A, Table I] In Table I, the 'No Fusion' and 'Early Fusion' rows report identical numbers on all three datasets; this is likely a typo or a copy-paste error and should be corrected or explained.
[Section IV.D] Figures 5 and 6 are referenced only with minimal captions and are not discussed in the body text with quantitative details; the authors should add explicit descriptions of what is plotted and which curves correspond to which setting.
[Section III.G] The loss formulation uses smooth L1 for regression and focal loss for both classification and prior knowledge supervision, but the weights alpha, beta, and gamma are not specified in the text. Please report the chosen values and whether they were tuned.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: GT supervision is training-time auxiliary signal; test-time predictions are learned from sensor data and evaluated on external benchmarks.

full rationale

Fast2comm uses ground-truth bounding boxes as training-time prior supervision, not as a substitute for the test-time prediction. The confidence map is produced by a learned generator from BEV features, and the GT-bbox binary map is used only to mask features for the training-time sharing; neither quantity is fed into the detector at inference as the final output. The paper explicitly decouples training and testing fusion in Section III.E, sharing only confidence features M at test, so the reported gains are not obtained by injecting GT boxes into the detector at inference. No load-bearing self-citation appears: the comparisons are against external methods such as Where2comm, MRCNet, Scope, and the benchmarks OPV2V, V2XSet, and DAIR-V2X are external. The train/test gap is a legitimate generalization concern, but it is not circularity: the test-time confidence map must still learn to localize from sensor data, and the ablations in Table II show the combined modules outperform the baseline without assuming the GT mask at test. Thus no step in the derivation reduces by construction to its own inputs.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central method relies on ground-truth bounding boxes as a training-time oracle, which is a strong assumption for real-world deployment. The confidence map threshold and loss weights are free hyperparameters not fully reported. The invented entities list is empty, as the paper introduces no new physical or mathematical objects beyond the modules.

free parameters (4)

Confidence map threshold t
Equation (1) uses a threshold t to binarize the confidence map; the value is not specified and must be chosen by hand, affecting which features are selected.
Loss weights alpha, beta, gamma
Equation (6) balances classification, regression, and prior-knowledge losses with weights; the values are not reported, but they are tuned hyperparameters.
Localization noise sigma_e = 0 to 0.5 meters
Gaussian noise standard deviation used to simulate localization errors; chosen as an experimental variable, not a learned parameter, but it affects the robustness evaluation.
Voxel size = 0.4 meters
PointPillars voxel dimensions set to 0.4m; a standard choice but still a hand-selected hyperparameter.

assumptions (3)

domain assumption Ground-truth labels are available for all agents during training.
The GT bbox-based feature selection and the prior loss rely on per-object ground truth for every agent in the training set, which may not hold in weakly-annotated or self-supervised settings.
domain assumption Feature maps from different agents can be aligned via coordinate transformation.
The method assumes a known calibration between agents, with the GT bbox projected into the ego coordinate system (Section III.C). The paper only simulates noise, not mis-calibration or unknown transformations.
ad hoc to paper The fusion network trained on concatenated confidence and GT-selected features remains effective when only confidence features are provided at test time.
This is the train-test decoupling assumption (Section III.E). The paper does not provide an architectural mechanism or empirical validation for this mismatch, making it a load-bearing but unverified premise.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Fast2comm:Collaborative perception combined with prior knowledge." pith.science (2026). https://pith.science/paper/DGEO337O

@misc{pith2026250500740,
  author       = {Pith},
  title        = {Pith review of: Fast2comm:Collaborative perception combined with prior knowledge},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DGEO337O}},
  note         = {Machine review of arXiv:2505.00740}
}

read the original abstract

Collaborative perception has the potential to significantly enhance perceptual accuracy through the sharing of complementary information among agents. However, real-world collaborative perception faces persistent challenges, particularly in balancing perception performance and bandwidth limitations, as well as coping with localization errors. To address these challenges, we propose Fast2comm, a prior knowledge-based collaborative perception framework. Specifically, (1)we propose a prior-supervised confidence feature generation method, that effectively distinguishes foreground from background by producing highly discriminative confidence features; (2)we propose GT Bounding Box-based spatial prior feature selection strategy to ensure that only the most informative prior-knowledge features are selected and shared, thereby minimizing background noise and optimizing bandwidth efficiency while enhancing adaptability to localization inaccuracies; (3)we decouple the feature fusion strategies between model training and testing phases, enabling dynamic bandwidth adaptation. To comprehensively validate our framework, we conduct extensive experiments on both real-world and simulated datasets. The results demonstrate the superior performance of our model and highlight the necessity of the proposed methods. Our code is available at https://github.com/Zhangzhengbin-TJ/Fast2comm.

Figures

Figures reproduced from arXiv: 2505.00740 by the authors.

**Figure 2.** The process of the proposed attention fusion module [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** The process of the proposed 4D bounding box [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 5.** Figure 5: Collaborative perception performance comparison with varying communication volume. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 6.** Figure 6: Robustness to localization error. Fast2comm outperforms baseline model and previous models. (a) (b) [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

**Figure 7.** Figure 7: Visualization of selected prior features. Brighter regions indicate the locations of the targets, while the blue boxes [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

**Figure 8.** Figure 8: Visualization of detection results. Green and red boxes represent ground-truth and detection results, respectively. [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 26 canonical work pages

[1]

V2vnet: Vehicle-to-vehicle communication for joint perception and prediction,

T.-H. Wang, S. Manivasagam, M. Liang, B. Yang, W. Zeng, and R. Ur- tasun, “V2vnet: Vehicle-to-vehicle communication for joint perception and prediction,” in Computer Vision ECCV 2020: 16th European Conference, Glasgow, UK, August 2328, 2020, Proceedings, Part II , 2020, pp. 605–621

work page 2020
[2]

V2x-vit: Vehicle-to-everything cooperative perception with vision trans- former,

R. Xu, H. Xiang, Z. Tu, X. Xia, M.-H. Yang, and J. Ma, “V2x-vit: Vehicle-to-everything cooperative perception with vision trans- former,” in Computer Vision ECCV 2022: 17th European Conference, Tel Aviv, Israel, October 2327, 2022, Proceedings. Berlin, Heidelberg: Springer-Verlag, 2022, pp. 107–124

work page 2022
[3]

Cooperative per- ception for 3d object detection in driving scenarios using infrastructure sensors,

E. Arnold, M. Dianati, R. de Temple, and S. Fallah, “Cooperative per- ception for 3d object detection in driving scenarios using infrastructure sensors,” IEEE Transactions on Intelligent Transportation Systems , vol. 23, no. 3, pp. 1852–1864, 2022

work page 2022
[4]

Learning distilled collaboration graph for multi-agent perception,

Y . Li, S. Ren, P. Wu, S. Chen, C. Feng, and W. Zhang, “Learning distilled collaboration graph for multi-agent perception,” in Advances in Neural Information Processing Systems , M. Ranzato, A. Beygelz- imer, Y . Dauphin, P. Liang, and J. W. Vaughan, Eds., vol. 34, 2021, pp. 29 541–29 552

work page 2021
[5]

A mechanism for scheduling multi robot intelligent warehouse system face with dynamic demand,

Z. Li, A. V . Barenji, J. Jiang, R. Y . Zhong, and G. Xu, “A mechanism for scheduling multi robot intelligent warehouse system face with dynamic demand,” J. Intell. Manuf., vol. 31, no. 2, pp. 469–480, Feb. 2020

work page 2020
[6]

When2com: Multi-agent perception via communication graph grouping,

Y .-C. Liu, J. Tian, N. Glaser, and Z. Kira, “When2com: Multi-agent perception via communication graph grouping,” in 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2020, pp. 4105–4114

work page 2020
[7]

Where2comm: communication-efficient collaborative perception via spatial confi- dence maps,

Y . Hu, S. Fang, Z. Lei, Y . Zhong, and S. Chen, “Where2comm: communication-efficient collaborative perception via spatial confi- dence maps,” in Proceedings of the 36st International Conference on Neural Information Processing Systems , ser. NIPS ’22, vol. 35, Red Hook, NY , USA, 2022, pp. 4874–4886

work page 2022
[8]

Multi-agent collaborative perception via motion-aware robust communication network,

S. Hong, Y . Liu, Z. Li, S. Li, and Y . He, “Multi-agent collaborative perception via motion-aware robust communication network,” in 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 15 301–15 310

work page 2024

Show all 26 references

[9]

Auction protocol for camera active control,

Y . Li, B. Bhanu, and W. Lin, “Auction protocol for camera active control,” in 2010 IEEE International Conference on Image Processing, 2010, pp. 4325–4328

2010
[10]

Multi-agent reinforcement learning: independent versus cooperative agents,

M. Tan, “Multi-agent reinforcement learning: independent versus cooperative agents,” in Proceedings of the Tenth International Confer- ence on International Conference on Machine Learning, ser. ICML’93, 1993, pp. 330–337

1993
[11]

Multi-agent graph-attention communication and teaming,

Y . Niu, R. Paleja, and M. Gombolay, “Multi-agent graph-attention communication and teaming,” in Proceedings of the 20th International Conference on Autonomous Agents and MultiAgent Systems , ser. AAMAS ’21. International Foundation for Autonomous Agents and Multiagent Systems...

2021
[12]

Communica- tion optimization for multi-agent reinforcement learning-based traffic control system with explainable protocol,

H. Wang, H. Wu, J. Lu, F. Tang, and M. L. D. Monache, “Communica- tion optimization for multi-agent reinforcement learning-based traffic control system with explainable protocol,” in 2023 IEEE 26th Interna- tional Conference on Intelligent Transportation Systems (ITSC) , 2023,...

2023
[13]

Opv2v: An open benchmark dataset and fusion pipeline for perception with vehicle-to-vehicle communication,

R. Xu, H. Xiang, X. Xia, X. Han, J. Li, and J. Ma, “Opv2v: An open benchmark dataset and fusion pipeline for perception with vehicle-to-vehicle communication,” in 2022 International Conference on Robotics and Automation (ICRA) , 2022, pp. 2583–2589

2022
[14]

Dair-v2x: A large-scale dataset for vehicle- infrastructure cooperative 3d object detection,

H. Yu, Y . Luo, M. Shu, Y . Huo, Z. Yang, Y . Shi, Z. Guo, H. Li, X. Hu, J. Yuan, and Z. Nie, “Dair-v2x: A large-scale dataset for vehicle- infrastructure cooperative 3d object detection,” in 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2022, pp...

2022
[15]

Scope: A synthetic multi-modal dataset for collective perception including physical-correct weather conditions,

J. Gamerdinger, S. Teufel, P. Schulz, S. Amann, J.-P. Kirchner, and O. Bringmann, “Scope: A synthetic multi-modal dataset for collective perception including physical-correct weather conditions,” in 2024 IEEE 27th International Conference on Intelligent Transportation Systems ...

2024
[16]

Dynamic feature sharing for cooperative perception from point clouds,

Z. Bai, G. Wu, M. J. Barth, Y . Liu, E. A. Sisbot, and K. Oguchi, “Dynamic feature sharing for cooperative perception from point clouds,” in 2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC) , 2023, pp. 3970–3976

2023
[17]

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 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019, pp. 12 689–12 697

2019
[18]

Ermvp: Communication-efficient and collaboration-robust multi-vehicle per- ception in challenging environments,

J. Zhang, K. Yang, Y . Wang, H. Wang, P. Sun, and L. Song, “Ermvp: Communication-efficient and collaboration-robust multi-vehicle per- ception in challenging environments,” in 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2024, pp. 12 575–12 584

2024
[19]

Attention is all you need,

A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, and I. Polosukhin, “Attention is all you need,” in Proceedings of the 31st International Conference on Neural Informa- tion Processing Systems , ser. NIPS’17, 2017, pp. 6000–6010

2017
[20]

Focal loss for dense object detection,

T.-Y . Lin, P. Goyal, R. Girshick, K. He, and P. Dollr, “Focal loss for dense object detection,” in 2017 IEEE International Conference on Computer Vision (ICCV) , 2017, pp. 2999–3007

2017
[21]

CARLA: An open urban driving simulator,

A. Dosovitskiy, G. Ros, F. Codevilla, A. Lopez, and V . Koltun, “CARLA: An open urban driving simulator,” in Proceedings of the 1st Annual Conference on Robot Learning , ser. Proceedings of Machine Learning Research, S. Levine, V . Vanhoucke, and K. Goldberg, Eds., vol. 78. PM...

2017
[22]

Pytorch: An imperative style, high- performance deep learning library,

A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, A. Desmaison, A. Kopf, E. Yang, Z. DeVito, M. Raison, A. Tejani, S. Chilamkurthy, B. Steiner, L. Fang, J. Bai, and S. Chintala, “Pytorch: An imperative style, high- p...

2019
[23]

Adam: A method for stochastic optimiza- tion,

D. P. Kingma and J. Ba, “Adam: A method for stochastic optimiza- tion,” in In International Conference on Learning Representations (ICLR), 12 2014

2014
[24]

Spatio-temporal domain awareness for multi- agent collaborative perception,

K. Yang, D. Yang, J. Zhang, M. Li, Y . Liu, J. Liu, H. Wang, P. Sun, and L. Song, “Spatio-temporal domain awareness for multi- agent collaborative perception,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) , October 2023, pp. 23 383–23 392

2023
[25]

Cobevt: Cooperative birds eye view semantic segmentation with sparse trans- formers,

R. Xu, Z. Tu, H. Xiang, W. Shao, B. Zhou, and J. Ma, “Cobevt: Cooperative birds eye view semantic segmentation with sparse trans- formers,” in Proceedings of The 6th Conference on Robot Learning , K. Liu, D. Kulic, and J. Ichnowski, Eds., vol. 205, 14–18 Dec 2023, pp. 989–1000

2023
[26]

How2comm: Communication-efficient and collaboration- pragmatic multi-agent perception,

D. Yang, K. Yang, Y . Wang, J. Liu, Z. Xu, R. Yin, P. Zhai, and L. Zhang, “How2comm: Communication-efficient and collaboration- pragmatic multi-agent perception,” in Advances in Neural Information Processing Systems, A. Oh, T. Naumann, A. Globerson, K. Saenko, M. Hardt, and S....

2023

Pith tools

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

[1] [1]

V2vnet: Vehicle-to-vehicle communication for joint perception and prediction,

T.-H. Wang, S. Manivasagam, M. Liang, B. Yang, W. Zeng, and R. Ur- tasun, “V2vnet: Vehicle-to-vehicle communication for joint perception and prediction,” in Computer Vision ECCV 2020: 16th European Conference, Glasgow, UK, August 2328, 2020, Proceedings, Part II , 2020, pp. 605–621

work page 2020

[2] [2]

V2x-vit: Vehicle-to-everything cooperative perception with vision trans- former,

R. Xu, H. Xiang, Z. Tu, X. Xia, M.-H. Yang, and J. Ma, “V2x-vit: Vehicle-to-everything cooperative perception with vision trans- former,” in Computer Vision ECCV 2022: 17th European Conference, Tel Aviv, Israel, October 2327, 2022, Proceedings. Berlin, Heidelberg: Springer-Verlag, 2022, pp. 107–124

work page 2022

[3] [3]

Cooperative per- ception for 3d object detection in driving scenarios using infrastructure sensors,

E. Arnold, M. Dianati, R. de Temple, and S. Fallah, “Cooperative per- ception for 3d object detection in driving scenarios using infrastructure sensors,” IEEE Transactions on Intelligent Transportation Systems , vol. 23, no. 3, pp. 1852–1864, 2022

work page 2022

[4] [4]

Learning distilled collaboration graph for multi-agent perception,

Y . Li, S. Ren, P. Wu, S. Chen, C. Feng, and W. Zhang, “Learning distilled collaboration graph for multi-agent perception,” in Advances in Neural Information Processing Systems , M. Ranzato, A. Beygelz- imer, Y . Dauphin, P. Liang, and J. W. Vaughan, Eds., vol. 34, 2021, pp. 29 541–29 552

work page 2021

[5] [5]

A mechanism for scheduling multi robot intelligent warehouse system face with dynamic demand,

Z. Li, A. V . Barenji, J. Jiang, R. Y . Zhong, and G. Xu, “A mechanism for scheduling multi robot intelligent warehouse system face with dynamic demand,” J. Intell. Manuf., vol. 31, no. 2, pp. 469–480, Feb. 2020

work page 2020

[6] [6]

When2com: Multi-agent perception via communication graph grouping,

Y .-C. Liu, J. Tian, N. Glaser, and Z. Kira, “When2com: Multi-agent perception via communication graph grouping,” in 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2020, pp. 4105–4114

work page 2020

[7] [7]

Where2comm: communication-efficient collaborative perception via spatial confi- dence maps,

Y . Hu, S. Fang, Z. Lei, Y . Zhong, and S. Chen, “Where2comm: communication-efficient collaborative perception via spatial confi- dence maps,” in Proceedings of the 36st International Conference on Neural Information Processing Systems , ser. NIPS ’22, vol. 35, Red Hook, NY , USA, 2022, pp. 4874–4886

work page 2022

[8] [8]

Multi-agent collaborative perception via motion-aware robust communication network,

S. Hong, Y . Liu, Z. Li, S. Li, and Y . He, “Multi-agent collaborative perception via motion-aware robust communication network,” in 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 15 301–15 310

work page 2024