Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

EffiComm: Bandwidth Efficient Multi Agent Communication

T0 review · 3 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read EffiComm claims two-stage pruning transmits less than 40% of prior art's data while keeping top 3D detection accuracy.

desk verdict Promising learned per-vehicle compression for V2X, but the headline numbers rest on a fuzzy byte metric and a GAT that needs its own signaling budget. read the letter →

arxiv 2507.19354 v1 pith:6ICMTIZJ submitted 2025-07-25 cs.CV cs.LGcs.RO

Melih Yazgan , Allen Xavier Arasan , J. Marius Zöllner This is my paper · ORCID
classification cs.CVcs.LGcs.RO
keywords collaborativeperceptionV2XcommunicationBEVfeaturecompressionselectivetransmissionadaptivegridreductionmixtureofexpertsgraphattentionnetwork3Dobjectdetection
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

EffiComm is an end-to-end pipeline for collaborative perception that claims to transmit less than 40% of the data required by earlier feature-sharing methods while matching or exceeding their 3D object-detection accuracy on the OPV2V benchmark. It works directly on bird's-eye-view feature maps, first pruning low-confidence spatial cells with a mask, then using a graph attention network to assign each vehicle a keep ratio that depends on its role and current network load. Surviving features are fused on the ego vehicle by a soft-gated mixture-of-experts attention layer. On OPV2V, the paper reports 0.843 AP@0.7 at roughly 1.9 MB per frame, against 0.83 at 4.9 MB for the nearest prior method. The claim matters because V2V bandwidth is the bottleneck that keeps collaborative perception from scaling to dense traffic.

What carries the argument

The load-bearing mechanism is a two-stage spatial-pruning pipeline on BEV features, followed by a mixture-of-experts attention fusion. Selective Transmission computes a class-confidence importance map and keeps only cells above a fixed threshold (inference) or a random top-$K$ subset (training). Adaptive Grid Reduction then runs a graph attention network over the vehicles, taking each vehicle's confidence-derived embedding, vehicle type, and the instantaneous transmission rate, to predict a keep ratio that is used to retain the top-$K_v$ highest-confidence cells. The resulting sparse maps are fused by a softmax-gated sum of several scaled dot-product attention experts. The argument is that the first stage removes useless background, the second coordinates competing vehicles, and the MoE head recovers capacity lost to pruning.

What would settle it

Instrument a full sparse-encoding version of EffiComm—index lists for kept cells, quantized feature values, per-vehicle keep ratios, and mask signaling—and transmit over a V2V channel emulator on the OPV2V test set. If the measured mean bytes per frame is not below 40% of a similarly instrumented Where2Comm run while keeping AP@0.7 within the reported margin, the paper's central efficiency claim is wrong.

Watch

Extended reading notes

Core claim

The central claim is that adaptive learned communication, not stronger sensors or bigger models, can close most of the accuracy gap between full-bandwidth feature sharing and bandwidth-limited sharing. EffiComm's two-stage reduction chooses which grid cells to keep, then how many each vehicle may keep, so that what is aired is concentrated where the detection head is most uncertain. The paper argues that the small accuracy cost of this sparsification is paid back by a soft-gated mixture-of-experts fusion head, which routes different spatial patterns to specialized attention experts and thereby integrates reduced features better than dense attention does. Empirically the paper reports the best accuracy-per-bit trade-off it compares against on OPV2V and competitive accuracy on the Culver City split, with communication cost measured as the $\log_2$ of the number of non-zero transmitted feature elements.

Load-bearing premise

The reported communication savings are computed from the $\log_2$ of the number of non-zero elements in the transmitted BEV features, and the paper does not specify how sparse indices, quantization, and per-vehicle mask overhead translate into actual bytes on the air.

Editorial extensions

If this is right

  • If the reported figures hold, collaborative perception systems can operate at V2V bandwidths below 2 MB/frame on OPV2V-like scenes without dropping below state-of-the-art detection accuracy.
  • The two-stage design separates content selection from per-agent budget assignment, implying the same framework can be dropped onto any BEV backbone, since the reduction operates on confidence maps rather than the raw sensor stream.
  • The transmission-rate feedback term in the keep-ratio equation gives a concrete mechanism for congestion-aware operation: as the instantaneous rate rises, every vehicle automatically prunes more aggressively.
  • Because the selective-transmission module is trained with random top-$K$ sparsity, the network is exposed to a range of sparsity patterns, which should make it robust to unseen communication budgets at deployment.

Reading between the lines

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

  • The headline '<40% of prior art' rests on a specific measurement convention ($\log_2$ of non-zero element count); a deployment planner should re-check the claim once sparse-index coding, quantization, and per-vehicle mask overhead are added to the byte count.
  • Because the MoE variant achieves lower bandwidth even with a higher average AGR keep ratio, the paper suggests the fusion head acts upstream as a feature regularizer; an ablation that separates the router's influence on the confidence map from its fusion capacity would isolate that effect.
  • The same keep-ratio feedback loop could be extended to include latency or packet-loss metrics, not just instantaneous rate, turning the graph network from a bandwidth coordinator into a general channel-adaptive scheduler.
  • The OPV2V benchmark uses simulated CARLA scenes; testing on real-world V2V traces would clarify whether the reported masks and keep ratios survive real sensor noise and pose error.
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 / 6 minor

Summary. The paper proposes EffiComm, an end-to-end framework for bandwidth-efficient V2V collaborative perception. It operates on BEV feature maps, applies a two-stage reduction pipeline (Selective Transmission, which prunes low-confidence spatial cells, and Adaptive Grid Reduction, which uses a graph attention network to assign per-vehicle keep ratios conditioned on role and network load), and fuses the surviving features with a soft-gated Mixture-of-Experts attention layer. On the OPV2V test set, the authors report AP@0.7 of 0.843 with a mean communication cost of 1.90 MB per frame, compared with 0.83 and 4.9 MB for Where2Comm, and claim that EffiComm transmits less than 40% of the data required by prior art while maintaining state-of-the-art detection accuracy. The paper also reports results on the Culver City dataset, an ablation of the MoE fusion head, and a variant trained with additional bandwidth and entropy losses.

Significance. If the communication-cost measurement were transparent and reproducible, this would be a solid contribution to collaborative perception: it shows that end-to-end training with only a detection loss can produce emergent sparsity that reduces transmitted features by roughly a factor of 2.5 relative to Where2Comm with a small accuracy gain, and the use of a GNN to output per-vehicle keep ratios is a plausible mechanism for load-aware compression. The evaluation is held-out on a standard benchmark, and the promised public release of code and weights would aid reproducibility. However, the central claim of the paper is a bandwidth claim, and the current manuscript does not define the byte-level measurement that supports it; the headline numbers in the abstract and the tables are not mutually consistent. The significance of the work can only be assessed after this measurement issue is resolved.

major comments (3)
  1. [§IV-A, Tables III-V] The central efficiency claim, that EffiComm 'transmits less than 40% of the data required by prior art,' is supported by the MB column in Table III, but no measurement procedure for these MB values is specified. Section IV-A defines the primary communication metric as log2 of the number of non-zero elements, which is not a byte count, and then states that raw bandwidth is 'directly measured' without describing the measurement: is it the number of retained feature elements times bytes per element, the size of a serialized sparse tensor, or an actual over-the-air V2V payload? No accounting is given for index coordinates, bitmasks, quantization, packetization, or the signaling needed for vehicles to know each other's masks and keep ratios. Because the headline ratio 1.90 MB vs 4.9 MB depends entirely on this definition, please specify the wire format or, failing that, an exact reproducible tensor-byte counting rule, and recompute the ratio under it. Also reconcile the abstract's 'approximately 1.5 MB per frame' with Table III's 1.90 MB and Table V's 1.37 MB; as written, the headline number matches no table entry.
  2. [Table IV footnote, Table I] The measurement definition is internally inconsistent. Table IV's footnote defines Comm as log2(non-zero elements × B) but leaves B undefined, while Tables I and II report Comm as a bare log2 count without any B factor. If B is bytes per element, then the log2 values in Table I are not directly comparable to the MB measurements in Table III, because each retained grid cell is a multi-channel feature vector rather than a single bit; if B is something else, it must be stated. As written, the reader cannot convert the reported log2 costs into MB, cannot check the claimed accuracy-per-bit trade-off, and cannot verify that the comparison to Where2Comm uses the same cost definition for all methods. Please use a single, clearly defined cost metric throughout, or report both the raw sparsity proxy and the byte-level measurement separately with their exact formulas.
  3. [§IV-A, Tables I-III] The comparison with prior methods may not be on equal footing because the manuscript does not state whether the same MB measurement procedure was applied to all baselines. Section IV-A says open-source weights were used for AttentiveFusion and FCooper and that Where2Comm was retrained, but it does not say whether the reported MB values for those methods include the same mask signaling, index overhead, or feature-channel counts as EffiComm. If the baselines' numbers were computed with different feature-map dimensions or without their own mask overhead, the '<40% of prior art' claim would be an artifact of the comparison protocol. Please specify the exact measurement pipeline for each method, including any compression or serialization steps.
minor comments (6)
  1. [§III-C] The sentence introducing node features says 'n_v = [e_hat_v, f_v, τ] where f_v is a binary flag and r represents the overall transmission rate,' but the variable r is introduced without definition and τ is not defined in that paragraph; please align the notation with Eq. (4).
  2. [§IV-D, Table III] The text states that EffiComm's maximum observed bandwidth is 2.66 MB and that Where2Comm's standard deviation is 0.35 MB, but Table III lists 2.44 MB and 0.37 MB respectively; please correct the text or the table.
  3. [Algorithm 1] Algorithm 1 line 7 says 'for each batch b do,' which is semantically wrong for a per-frame inference procedure; this should be per scene or per frame, and the variables F_b, M_b, and τ_b should be defined with respect to vehicles and time steps.
  4. [Tables I-II] The table formatting is inconsistent: on Culver City, EffiComm does not achieve the best AP, yet the text says it is 'competitive'; this is acceptable, but the tables should be explicit about which entries are best and the text should be precise about where EffiComm is not state-of-the-art.
  5. [References] Reference [5] has a malformed author list ('J. L. J. M. Runsheng Xu Hao Xiang, Xin Xia Xu Han'); please correct it to the OPV2V author list.
  6. [Figures 4-5] Figure 4's caption does not define the colors used for the three keep-ratio distributions, and Figure 5's six panels are not labeled in the caption; please add legends and panel labels so the reader can follow the discussion in Sections IV-E and IV-F.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: EffiComm's bandwidth savings are measured on held-out data, not derived from a fitted constant.

full rationale

The paper's central claim is an empirical comparison on the OPV2V benchmark: EffiComm reports 0.843 AP@0.7 with 1.90 MB/frame (Table III) versus 0.83 and 4.9 MB for Where2Comm, and the headline '<40% of prior art' follows arithmetically from these measurements. The main model is trained end-to-end with the detection loss only (Section IV, paragraph before IV-A), so the bandwidth reduction is an emergent property evaluated on a held-out test set rather than a quantity fitted by construction. The only loss that directly minimizes the reported communication metric appears in the ablation labeled 'Impact of Bandwidth and Entropy Loss Terms' (Section IV-F, Table V), where L_bw is explicitly introduced as an additional training term; this is presented as an ablation, not as the source of the main efficiency claim. The communication metric 'Comm = log2(non-zero elements)' is defined in Section IV-A as a proxy following prior work, and the paper also reports raw MB statistics; while the MB measurement procedure is under-specified and the abstract's '~1.5 MB' is inconsistent with Table III's 1.90 MB, this is a measurement-fidelity/correctness issue, not circularity. The self-citations ([3], [25]) are survey references and are not load-bearing for any technical derivation. No equation in the paper reduces to its own inputs, no fitted parameter is renamed as a prediction, and no uniqueness theorem or prior-work ansatz is invoked to force the result. The central derivation chain is therefore self-contained with respect to circularity concerns.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the benchmarks being representative, the confidence map being a valid utility signal, and the non-zero-element count being a faithful stand-in for real bandwidth. The free parameters (threshold mu, base keep ratios, and the Eq. 4 coefficients) are hand-chosen and directly modulate the reported bandwidth-accuracy trade-off. No new entities are invented.

free parameters (6)
  • ST confidence threshold mu = 0.01
    Fixed at inference in Section IV; binarizes the sigmoid-normalized confidence map. It directly controls how many cells survive the first pruning stage, and the paper provides no sensitivity analysis.
  • Ego base keep ratio k_ego = 0.9
    Hand-set in Section III-C; enters Eq. 4 and scales the ego vehicle's survival fraction. The value is chosen to keep high fidelity but is not derived.
  • Remote base keep ratio k_remote = 0.5
    Hand-set in Section III-C; enters Eq. 4 and scales remote vehicles' survival fractions. Directly trades accuracy against bandwidth.
  • Keep-ratio formula coefficients = 0.5, 0.5, 0.7, 0.3
    Coefficients in Eq. 4: (0.5+0.5 alpha) and (0.7+0.3(1-tau)). They define how GAT outputs and network load modulate keep ratios, without a stated justification.
  • Bandwidth loss weight lambda_bw = 0.05
    Used in the +AL variant (Section IV-F) to penalize the proportion of non-zero transmitted features; this loss directly targets the reported Comm metric.
  • Entropy regularization weight mu_ent = 1e-4
    Used in the +AL variant to regularize MoE gating expert utilization; reported in Section IV-F.
assumptions (4)
  • domain assumption OPV2V and its Culver City subset are representative of real V2V cooperative perception tasks.
    All accuracy and bandwidth conclusions are drawn from these datasets (Section IV); if the simulation-to-real gap is large, the trade-off could differ in deployment.
  • domain assumption The detection head's class-confidence map is a reliable proxy for the spatial utility of BEV features.
    Both ST and AGR select cells using confidence scores (Eq. 1 and Eq. 5-6); if confidence does not track task-relevant information, pruning will discard useful features.
  • ad hoc to paper The number of non-zero feature elements is a valid proxy for over-the-air bandwidth.
    The Comm metric is defined in Section IV-A as log2 of non-zero count; actual byte cost depends on index coding, quantization, and packetization, which are not modeled.
  • domain assumption Random top-K dropout during training transfers to threshold-based inference.
    Section III-B introduces the train/inference mismatch; the paper treats it as regularization without supporting evidence.

how reviews work

0 comments
Cite this review

Pith. "Pith review of EffiComm: Bandwidth Efficient Multi Agent Communication." pith.science (2026). https://pith.science/paper/6ICMTIZJ

@misc{pith2026250719354,
  author       = {Pith},
  title        = {Pith review of: EffiComm: Bandwidth Efficient Multi Agent Communication},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6ICMTIZJ}},
  note         = {Machine review of arXiv:2507.19354}
}
read the original abstract

Collaborative perception allows connected vehicles to exchange sensor information and overcome each vehicle's blind spots. Yet transmitting raw point clouds or full feature maps overwhelms Vehicle-to-Vehicle (V2V) communications, causing latency and scalability problems. We introduce EffiComm, an end-to-end framework that transmits less than 40% of the data required by prior art while maintaining state-of-the-art 3D object detection accuracy. EffiComm operates on Bird's-Eye-View (BEV) feature maps from any modality and applies a two-stage reduction pipeline: (1) Selective Transmission (ST) prunes low-utility regions with a confidence mask; (2) Adaptive Grid Reduction (AGR) uses a Graph Neural Network (GNN) to assign vehicle-specific keep ratios according to role and network load. The remaining features are fused with a soft-gated Mixture-of-Experts (MoE) attention layer, offering greater capacity and specialization for effective feature integration. On the OPV2V benchmark, EffiComm reaches 0.84 mAP@0.7 while sending only an average of approximately 1.5 MB per frame, outperforming previous methods on the accuracy-per-bit curve. These results highlight the value of adaptive, learned communication for scalable Vehicle-to-Everything (V2X) perception.

Figures

Figures reproduced from arXiv: 2507.19354 by the authors.

Figure 1
Figure 1. End-to-end pipeline illustrating feature extraction, the two-stage reduction process (Selective Transmission and [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Internal components of the adaptive reduction stage. A GNN processes vehicle features (derived from confidence [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Comparison of feature selection between Where2comm and EffiComm. EffiComm selects fewer but more relevant features from other agents to the ego vehicle. C. Performance Comparison As demonstrated in Table I, our proposed EffiComm approach achieves better detection performance (IoU@0.5: 0.92, IoU@0.7: 0.843) than state-of-the-art methods like AttentiveFusion and Where2Comm. Also, EffiComm sig￾nificantly reduces the co… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Distribution of predicted keep-ratios for remote [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Interpretability of MoE fusion. Upper-left: t-SNE of routed tokens coloured by expert ID showing three dis￾tinct clusters. Upper-right: density contours highlight cluster boundaries. Lower-left: centroid-distance heatmap (0.8–5.7) confirms separation. Lower-right: aver…

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. SlimComm: Doppler-Guided Sparse Queries for Bandwidth-Efficient Cooperative 3-D Perception

    cs.CV 2025-08 conditional novelty 7.0 of 10

    Doppler-guided sparse queries with two-stage offsets let CAVs share about 10 percent of the BEV feature data at matching detection accuracy.

Reference graph

Works this paper leans on

25 extracted references · 20 canonical work pages · cited by 1 Pith paper

  1. [1]

    Deep continuous fusion for multi-sensor 3d object detection,

    M. Liang, B. Yang, S. Wang, and R. Urtasun, “Deep continuous fusion for multi-sensor 3d object detection,” in ECCV, 2018

    work page 2018

  2. [2]

    Robust real-time multi-vehicle collaboration on asynchronous sensors,

    Q. Zhang, X. Zhang, R. Zhu, F. Bai, M. Naserian, and Z. M. Mao, “Robust real-time multi-vehicle collaboration on asynchronous sensors,” ser. ACM MobiCom, 2023

    work page 2023

  3. [3]

    Real- world problems in collaborative perception: A categorized review of intermediate fusion methods,

    M. Yazgan, T. Graf, M. Liu, T. Fleck, and J. M. Z ¨ollner, “Real- world problems in collaborative perception: A categorized review of intermediate fusion methods,” IEEE IV, 2024

    work page 2024

  4. [4]

    Evolutionary v2x technolo- gies toward the internet of vehicles: Challenges and opportunities,

    H. Zhou, W. Xu, J. Chen, and W. Wang, “Evolutionary v2x technolo- gies toward the internet of vehicles: Challenges and opportunities,” Proceedings of the IEEE , 2020

    work page 2020

  5. [5]

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

    J. L. J. M. Runsheng Xu Hao Xiang, Xin Xia Xu Han, “Opv2v: An open benchmark dataset and fusion pipeline for perception with vehicle-to-vehicle communication,” in 2022 IEEE International Con- ference on Robotics and Automation (ICRA) , 2022

    work page 2022

  6. [6]

    F-cooper: Feature based cooperative perception for autonomous vehicle edge computing using 3d point clouds,

    Q. Chen, S. Tang, Q. Yang, and S. Fu, “F-cooper: Feature based cooperative perception for autonomous vehicle edge computing using 3d point clouds,” in ACM EdgeSys, 2019

    work page 2019

  7. [7]

    Discograph: Learning distilled collaboration graph for multi-agent perception,

    Y . Li, S. Ren, P. Wu, S. Chen, C. Feng, and W. Zhang, “Discograph: Learning distilled collaboration graph for multi-agent perception,” in NeurIPS, 2021

    work page 2021

  8. [8]

    Vehicle–infrastructure cooperative 3d object detection via feature flow prediction,

    H. Yu, Y . Tang, E. Xie, J. Mao, J. Yuan, P. Luo, and Z. Nie, “Vehicle–infrastructure cooperative 3d object detection via feature flow prediction,” in NeurIPS, 2023

    work page 2023

Show all 25 references
  1. [9]

    Slim-fcp: Lightweight feature-based cooperative perception for connected au- tomated vehicles,

    J. Guo, D. Carrillo, Q. Chen, Q. Yang, and S.-F. Fu, “Slim-fcp: Lightweight feature-based cooperative perception for connected au- tomated vehicles,” in IEEE ICCPS, 2022

    2022

  2. [10]

    Vimi: Vehicle–infrastructure multi-view intermediate fusion for camera-based 3d object detection,

    Z. Wang, S. Fan, X. Huo, T. Xu et al., “Vimi: Vehicle–infrastructure multi-view intermediate fusion for camera-based 3d object detection,” arXiv preprint arXiv:2303.10975 , 2023

    2023 arXiv

  3. [11]

    Scalable feature compres- sion for edge-assisted object detection over time-varying networks,

    Z. Yuan, S. Garg, E. Erkip, and Y . Wang, “Scalable feature compres- sion for edge-assisted object detection over time-varying networks,” in ICMLSys Workshop on Resource-Constrained Learning in Wireless Networks, 2023

    2023

  4. [12]

    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 Proceedings of the IEEE/CVF Conference on computer vision and pattern recognition , 2020, pp. 4106–4115

    2020

  5. [13]

    Select2col: Leveraging spatial-temporal importance of semantic information for efficient collaborative perception,

    Y . Liu, Q. Huang, R. Li, X. Chen, Z. Zhao, S. Zhao, Y . Zhu, and H. Zhang, “Select2col: Leveraging spatial-temporal importance of semantic information for efficient collaborative perception,” IEEE Transactions on Vehicular Technology , vol. 73, no. 9, pp. 12 556– 12 568, 2024

    2024

  6. [14]

    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 NeurIPS, 2022

    2022

  7. [15]

    Bm2cp: Efficient collaborative perception with lidar-camera modalities,

    B. Zhao, W. Zhang, and Z. Zou, “Bm2cp: Efficient collaborative perception with lidar-camera modalities,” in CoRL, 2023

    2023

  8. [16]

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

    W. Xu, X. Cai, X. Zhang, L. Lin, C. Liu, W. Ouyang, and L. You, “V2x-vit: Vehicle-to-everything cooperative perception with vision transformer,” in European Conference on Computer Vision. Springer, 2022, pp. 106–123

    2022

  9. [17]

    Edgecooper: Cooperative perception with edge attention for autonomous driving,

    M. Ye, Y . Yao, J. Xu, C. Shen, and C. An, “Edgecooper: Cooperative perception with edge attention for autonomous driving,” in 2023 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2023, pp. 11 681–11 687

    2023

  10. [18]

    Outrageously large neural networks: The sparsely- gated mixture-of-experts layer,

    N. e. a. Shazeer, “Outrageously large neural networks: The sparsely- gated mixture-of-experts layer,” in ICLR, 2017

    2017

  11. [19]

    Scaling vision with sparse mixture of experts,

    C. Riquelme, J. Puigcerver, B. Mustafa, M. Neumann, R. Jenatton, A. S. Pinto, D. Keysers, and N. Houlsby, “Scaling vision with sparse mixture of experts,” ser. NIPS, 2021

    2021

  12. [20]

    Pointpillars: Fast encoders for object detection from point clouds,

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

    2019

  13. [21]

    Insightdrive: Insight scene representation for end-to-end autonomous driving,

    R. Song, X. Guo, H. Wu, Q. Wei, and L. Chen, “Insightdrive: Insight scene representation for end-to-end autonomous driving,” 2025

    2025

  14. [22]

    Rate distortion theory: A mathematical basis for data compression,

    L. Davisson, “Rate distortion theory: A mathematical basis for data compression,” IEEE Transactions on Communications , vol. 20, no. 6, pp. 1202–1202, 1972

    1972

  15. [23]

    Opencda: an open cooperative driving automation framework integrated with co-simulation,

    R. Xu, Y . Guo, X. Han, X. Xia, H. Xiang, and J. Ma, “Opencda: an open cooperative driving automation framework integrated with co-simulation,” in 2021 IEEE International Intelligent Transportation Systems Conference (ITSC) . IEEE, 2021, pp. 1155–1162

    2021

  16. [24]

    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 , 2017, pp. 1–16

    2017

  17. [25]

    Collabo- rative perception datasets in autonomous driving: A survey,

    M. Yazgan, M. V . Akkanapragada, and J. Marius Z ¨ollner, “Collabo- rative perception datasets in autonomous driving: A survey,” in 2024 IEEE Intelligent Vehicles Symposium (IV) , 2024, pp. 2269–2276

    2024

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.