Pith. sign in

REVIEW 3 major objections 5 minor 52 references

Generate Realistic Test Scenes for V2X Communication Systems

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read V2XGen claims that automatically transforming real multi-vehicle LiDAR scenes — inserting, deleting, scaling, rotating, and translating objects while keeping every agent's viewpoint consistent — yields realistic V2X test scenes that…

desk verdict A plausible, useful V2X test-generation tool whose central claim—cross-agent perspective consistency—is asserted but never directly measured; needs one focused revision. read the letter →

arxiv 2506.07419 v1 pith:AWQAYNU2 submitted 2025-06-09 cs.SE

An Guo , Xinyu Gao , Chunrong Fang , Haoxiang Tian , Weisong Sun , Yanzhou Mu , Shuncheng Tang , Lei Ma
show 1 more author
Zhenyu Chen
This is my paper
classification cs.SE
keywords V2XcooperativeperceptionautomatedtestgenerationmetamorphictestingLiDARpointcloudsocclusionerrorlong-rangefitness-guidedautonomousdriving
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

V2XGen is an automated tool for generating test scenes for Vehicle-to-Everything cooperative perception systems, in which vehicles share LiDAR point clouds to detect objects that a single car would miss due to occlusion or distance. The paper's claim is that by transforming real multi-vehicle scenes with five operators (insertion, deletion, scaling, rotation, and translation) while preserving viewpoint consistency through coordinate transforms and virtual LiDAR rendering, V2XGen produces realistic test data that exposes two V2X-specific failure classes: occlusion perception errors and long-range perception errors. A fitness-guided strategy scores candidate scenes by their predicted ability to trigger those failures and keeps the highest-scoring ones, which the authors show finds more errors than random selection or the existing CooTest approach. The paper further claims that retraining cooperative perception systems on the generated scenes improves average detection precision while reducing occlusion and long-range errors. If correct, this replaces costly manual collection and labeling with an automated loop for testing and improving cooperative driving perception.

What carries the argument

Three mechanisms carry the argument. The multi-agent perspective transformation (Equation 3) maps world coordinates into each participant's local coordinate system so that any inserted or moved entity occupies the same physical location in every agent's point cloud. A set of virtual LiDAR sensors, initialized to the real sensor's configuration (Velodyne VLP-32), uses ray-casting to render inserted entities, delete occluded points, and complete ground and objects behind deleted entities, so scenes follow physical occlusion laws. The fitness metric (Equation 6) combines an occlusion-error score and a long-range-error score that weight missed objects by their occlusion rate and distance, guiding scene selection toward cases where the ego vehicle needs help most; metamorphic relations (input-output properties that must still hold after transformation) built on insertion and deletion supply the test oracle, so labels come from the transformation itself rather than manual annotation.

What would settle it

Have annotators label a random sample of V2XGen-generated scenes: if the human-labeled bounding boxes of inserted objects fall below the IoU 0.5 threshold against V2XGen's metadata-derived labels in any cooperating vehicle's view, the perspective-consistency assumption breaks. A sharper test is to perturb the extrinsic pose matrices by the noise levels documented in V2V4Real and measure how much cross-view object alignment and FRD realism degrade; if realistic noise already misaligns the rendered object across agents, the residual perception errors after retraining are explained by the assumption failing rather than by system weaknesses.

Watch

Extended reading notes

Core claim

The central claim is that V2XGen generates realistic V2X test scenes, detects erroneous behaviors in cooperative perception systems operating under different fusion schemes, and that retraining the systems under test with the generated scenes enhances average detection precision while reducing occlusion and long-range perception errors. The realism is carried by perspective consistency: each transformation operator renders an entity through per-agent virtual LiDARs and intersection with background geometry, so a manipulated object appears from every cooperative vehicle's viewpoint as it would in the world. The targeted errors are formalized as occlusion perception error (the system misses an object that is occluded from the ego vehicle but visible to a cooperative vehicle) and long-range perception error (the system misses an object beyond the ego vehicle's range even though a cooperative vehicle could assist). The paper supports the claim with FRD realism scores and a user study favoring V2XGen over single-agent baselines, higher occlusion and long-range error rates than CooTest and unguided generation on six systems, and before/after retraining AP improvements across all six systems.

Load-bearing premise

The load-bearing premise is that the coordinate transformations between vehicles are accurate enough for a rendered object to land in the same physical position in every vehicle's point cloud; the paper itself reports in Section 5.3 that V2V4Real has a developer-confirmed positioning error that may adversely affect retraining.

Editorial extensions

If this is right

  • V2X test data can be generated and labeled automatically from real seed scenes, eliminating manual data collection and annotation for the covered scenarios.
  • Fitness-guided selection concentrates the test budget on occlusion and long-range cases, detecting more perception failures per generated test than random or CooTest-guided selection.
  • Retraining cooperative perception systems on generated scenes improves detection precision on both transformed and original test sets, with reported average occlusion-error reductions of 54.9% and long-range-error reductions of 30.2%.
  • Because the operators work on any multi-view LiDAR scene with pose annotations, the approach extends across early, late, and intermediate fusion architectures.

Reading between the lines

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

  • The perspective-consistency machinery could transfer to vehicle-infrastructure (V2I) cooperative perception, where one sensor has a fixed elevated pose; the insertion operator's validity checks would need to handle different road-surface and occlusion geometry.
  • The fitness metric is open-ended: adding new score terms (e.g., for misclassification of rescaled vehicle shapes or for weather-degraded views) could steer the same rendering pipeline toward additional failure classes without retraining the generator.
  • The measured realism and retraining gains are tied to V2V4Real's Velodyne VLP-32 configuration; re-initializing the virtual LiDAR to other sensor models (solid-state, different beam counts) would test whether the rendering approach transfers.
  • Because the fitness scores favor scenes where the ego vehicle is occluded or short-ranged, a retrained model's behavior on ordinary balanced traffic may shift; measuring per-scenario AP distributions, not just averages, would reveal such trade-offs.
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

3 major / 5 minor

Summary. The paper presents V2XGen, an automated test-scene generation tool for V2X cooperative perception systems. V2XGen applies five transformation operators (insertion, deletion, scale, rotation, translation) to multi-agent LiDAR scenes and claims to maintain perspective consistency across agents through a world-to-vehicle coordinate transformation (Eq. 3). Generated scenes are selected through a fitness-guided strategy (Eqs. 4-6) that targets occluded and long-range objects. The authors evaluate the tool on six cooperative detection systems with different fusion schemes using the V2V4Real dataset. They report (i) realism via FRD and a 22-participant user study; (ii) error-detection effectiveness against CooTest and a non-guided variant; and (iii) improvements in average precision and reduced perception errors after retraining.

Significance. If the results hold, V2XGen would be a useful contribution to testing and robustness improvement for cooperative perception systems. The paper addresses a real gap: single-agent test generators do not automatically produce multi-view consistent scenes, and the existing CooTest baseline does not support object-level manipulation. The tool and generated scenes are released, and the evaluation covers early, late, and intermediate fusion systems. The main limitations are that the central realism claim is not directly validated for cross-agent consistency, and the efficiency claims are partly measured with the same error definitions used in the selection objective. These issues are addressable with additional validation and statistical rigor, but they are load-bearing for the paper's central claims.

major comments (3)
  1. [Section 3.2.1 and Section 5.1] The paper's central differentiator is perspective consistency across agents, but this property is never directly measured. Equation (3) maps a single world-space point to each agent's local LiDAR frame, so if the seed dataset's extrinsic transforms are not accurate, an inserted object will appear at different physical locations in different agents' point clouds and the generated scene will not be a valid V2X scene. The realism metrics used in RQ1 do not detect this failure: FRD in Eq. (10) is computed per agent on RangeNet++ activations, and the user study asks participants to judge ego and cooperative data frames separately. Moreover, Section 5.3 acknowledges a developer-confirmed positioning error in the V2V4Real dataset that may adversely affect retraining. The authors should quantify cross-agent alignment directly, for example by rendering a known object into multiple views, transforming the resulting point clouds back to world coordinates, and measuring the dispersion of the object center, and they should report how pose error or misalignment tolerance affects scene realism and retraining results.
  2. [Section 3.3 and Section 5.2 (Eqs. 4-6; Tables 2-3)] The fitness-guided selection objective and the RQ2 evaluation metrics are aligned by construction. The scores FOP and FLP in Eqs. (4) and (5) include the indicator IGT(Bgt, Bp), which is true when the system under test fails to detect a ground-truth object, and Algorithm 1 retains only high-fitness scenes. Tables 2 and 3 then count occlusion and long-range perception errors on the retained set, so higher error counts compared with random retention are a partial reflection of the optimization objective rather than an independent demonstration of testing effectiveness. To support the claim of improved testing efficiency, the authors should use an independent evaluation, for example measuring detection of previously unknown failures on a held-out set of transformed scenes, or evaluating whether scenes selected using one system expose errors in a different system without re-optimizing the fitness function.
  3. [Section 5 (Tables 1-4)] All quantitative results are reported as averages over three or five runs without standard deviations, confidence intervals, or significance tests. Since the generation process involves random operator selection, random seeds, and randomly selected subsets, the reader cannot determine whether the reported differences are reliable. The only statistical test reported is the Wilcoxon rank-sum test for the user study in Section 5.1. The authors should report per-run results or variance for Tables 1-4 and apply appropriate significance tests or effect-size measures; without this, claims such as the 3.3%-15.5% retraining improvements in Table 4 cannot be distinguished from random variation.
minor comments (5)
  1. [Throughout] The tool name is rendered inconsistently as V2XGen and V2XGEN; please unify the notation.
  2. [Section 4.4] The SAGen baseline is described as single-agent operators combined with Eq. (3) to enforce cross-view consistency; if this is the case, the comparison isolates operator-level fidelity while still assuming the same coordinate transforms, so the baseline does not test the full multi-agent consistency claim. Please clarify what is being ablated.
  3. [Section 3.4] The metamorphic relations in Eqs. (7) and (8) are written with notation such as zeta{CPJsK union GTins, CPJsigma(s, eins)K}, but the intended equality criterion is only described in words as soft equality derived from AP; a precise formal definition would improve reproducibility.
  4. [Section 5.1] The FRD values in Table 1 are reported without units or an explanation of the scale, making it difficult to interpret whether a reduction from 194.3 to 184.2 is practically meaningful; please state the range of the metric and report the real-vs-real baseline if available.
  5. [Section 5.2] Table 2 reports absolute OE and LE counts without giving the total number of objects or scenes in the retained test sets; the error rates in Table 3 are more interpretable, but the denominator should be stated explicitly for both tables.

Circularity Check

1 steps flagged · score 6.0 of 10

RQ2's error-detection advantage is partly self-definitional: the fitness metric that selects scenes is built from the same occlusion and long-range error conditions later counted, so the gain over random retention is guaranteed by the selection rule rather than independently established.

  1. self definitional [Section 3.3 (Eqs. 4-6 and Algorithm 1) vs. Section 5.2 (Tables 2-3)]
    "Finally, our fitness metric can be expressed as a weighted sum of the two scores: F itness(m) =α ∗ FOP + β ∗ FLP (6) ... if Spc.F itnesspri > TD[−1].F itnesspri then TD[−1] =Spc; TD.SortByF itness() ... Compared to the V2XG EN N and CooTest guidance strategy, the proposed fitness-guided V2X scene generation method can find more occlusion and long-range perception errors in all configurations."

    FOP (Eq. 4) and FLP (Eq. 5) are defined with the exact ingredients later counted as OE and LE: IGT indicates the system missed the ground-truth box (IoU ≤ γ), Occego weights ego occlusion, and the distance terms weight long-range failure. Algorithm 1 keeps only the scenes with the largest Fitnesspri. Therefore Tables 2 and 3 compare V2XGen against random retention on the very objective used for selection; higher OE/LE counts and rates are a mathematical consequence of picking the top-fitness scenes, not an independent empirical finding that the generated data are more error-revealing. The lower AP reported for V2XGen test sets is likewise the direct result of selecting scenes in which IGT=1 for many boxes.

full rationale

The paper's main non-circular content is substantial: RQ1 evaluates realism with an external FRD metric on RangeNet++ activations and a human preference study, and RQ3 shows AP gains on the original V2V4Real test scenes after retraining, which is an out-of-distribution check not encoded in the fitness function. The circular portion is confined to RQ2's headline that the fitness-guided strategy finds more occlusion and long-range perception errors than random retention. Because FOP/FLP are weighted versions of those very error conditions and Algorithm 1 retains top-fitness scenes, the observed OE/LE advantage is partially guaranteed by construction. This is not a case of a fitted parameter renamed as a prediction, nor does the paper rely on load-bearing self-citations: CooTest and LiRTest are used as baselines/prior metamorphic-testing practice, not to justify the central claim. The acknowledged V2V4Real positioning error (Sec. 5.3) is a correctness/validity threat to perspective consistency, not a circularity, because the paper does not define realism in terms of the extrinsic transform. Overall, one central evaluation claim reduces by construction, giving partial circularity despite independent realism and retraining evidence.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The central claims rest on assumptions about dataset pose accuracy, LiDAR simulation fidelity, semantic segmentation correctness, and the validity of the metamorphic oracle. The paper identifies the pose error itself, which makes that assumption particularly fragile. No new physical entities are introduced; the fitness metric is an optimization objective rather than an explanation mechanism.

free parameters (5)
  • alpha, beta (fitness weights) = 0.5, 0.5
    Hand-set in Section 4.1 with no sensitivity analysis; the RQ2 comparison depends on this weighting of occlusion versus long-range errors.
  • k (long-range distance threshold) = 50 m
    Set in Section 4.1 according to V2V4Real's distance divisions; it defines which detection failures count as long-range perception errors in Equations (2) and (5).
  • N (maximum entity manipulations per scene) = 3
    Set in Section 4.1; it controls the mutation budget, and no ablation over N is provided.
  • SC and RO parameter ranges = sx, sy, sz in [0.9, 1.1]; rot in [-30, -5] or [5, 30] degrees
    Chosen as conservative realism ranges in Section 3.2.4; these bounds affect which scenes are generated and how realistic they appear.
  • Retention ratio for fitness-guided selection = 10% and 15% of transformed data
    Experimental choices in RQ2 that determine the size and composition of test sets; results may vary with the retention ratio.
assumptions (5)
  • domain assumption The extrinsic coordinate transforms between the ego vehicle and cooperative vehicles in V2V4Real are accurate enough to place rendered objects consistently in every agent's frame.
    Used in Equation (3) and throughout Section 3.2 to maintain perspective consistency; Section 5.3 concedes the dataset's developer confirmed a positioning error that may affect retraining.
  • domain assumption A virtual LiDAR configured like the Velodyne VLP-32 faithfully simulates real LiDAR rays, including occlusion and ground completion.
    The realism of inserted and deleted point clouds depends on this simulation; configuration is given in Section 4.1 and the operators are described in Sections 3.2.2 to 3.2.4.
  • domain assumption The metamorphic relations MR1 and MR2 hold: inserting or deleting an object should not change the correct perception results for other objects.
    Used as the test oracle in Section 3.4; if real perception systems legitimately change detections under these transformations, some reported failures may be false positives.
  • domain assumption The semantic segmentation model identifies road surfaces correctly in every agent's point cloud.
    Valid insertion locations are computed by separating road points in Section 3.2.2, Step 1; segmentation errors would place objects off-road or in invalid positions.
  • domain assumption Ground-truth bounding boxes in the V2V4Real dataset are accurate.
    GT labels are used to delete objects, generate new labels, and compute occlusion and long-range errors and average precision; the paper itself notes a positioning error in the dataset in Section 5.3.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Generate Realistic Test Scenes for V2X Communication Systems." pith.science (2026). https://pith.science/paper/AWQAYNU2

@misc{pith2026250607419,
  author       = {Pith},
  title        = {Pith review of: Generate Realistic Test Scenes for V2X Communication Systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AWQAYNU2}},
  note         = {Machine review of arXiv:2506.07419}
}
read the original abstract

Accurately perceiving complex driving environments is essential for ensuring the safe operation of autonomous vehicles. With the tremendous progress in deep learning and communication technologies, cooperative perception with Vehicle-to-Everything (V2X) technologies has emerged as a solution to overcome the limitations of single-agent perception systems in perceiving distant objects and occlusions. Despite the considerable advancements, V2X cooperative perception systems require thorough testing and continuous enhancement of system performance. Given that V2X driving scenes entail intricate communications with multiple vehicles across various geographic locations, creating V2X test scenes for these systems poses a significant challenge. Moreover, current testing methodologies rely on manual data collection and labeling, which are both time-consuming and costly. In this paper, we design and implement V2XGen, an automated testing generation tool for V2X cooperative perception systems. V2XGen utilizes a high-fidelity approach to generate realistic cooperative object instances and strategically place them within the background data in crucial positions. Furthermore, V2XGen adopts a fitness-guided V2X scene generation strategy for the transformed scene generation process and improves testing efficiency. We conduct experiments on V2XGen using multiple cooperative perception systems with different fusion schemes to assess its performance on various tasks. The experimental results demonstrate that V2XGen is capable of generating realistic test scenes and effectively detecting erroneous behaviors in different V2X-oriented driving conditions. Furthermore, the results validate that retraining systems under test with the generated scenes can enhance average detection precision while reducing occlusion and long-range perception errors.

Figures

Figures reproduced from arXiv: 2506.07419 by the authors.

Figure 1
Figure 1. The architecture of cooperative perception systems in au [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Motivating examples illustrating the core functionalities of V2X [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The workflow of V2XGEN [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Virtual LiDAR equipped by the V2X cooperative perception [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Subfigures (a) and (b) illustrate the visualization effects of [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Participants’ choices over different data synthesis methods in [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

52 extracted references · 28 canonical work pages

  1. [1]

    Autonomous driving system: A comprehensive survey,

    J. Zhao, W. Zhao, B. Deng, Z. Wang, F. Zhang, W. Zheng, W. Cao, J. Nan, Y. Lian, and A. F. Burke, “Autonomous driving system: A comprehensive survey,” Expert Syst. Appl. , vol. 242, p. 122836,

    work page

  2. [2]

    Lidar for autonomous driving: The principles, challenges, and trends for automotive lidar and per- ception systems,

    Y. Li and J. Ibanez-Guzman, “Lidar for autonomous driving: The principles, challenges, and trends for automotive lidar and per- ception systems,” IEEE Signal Processing Magazine , vol. 37, no. 4, pp. 50–61, 2020

    work page 2020

  3. [3]

    Deep learning-based perception systems for autonomous driving: A comprehensive survey,

    L.-H. Wen and K.-H. Jo, “Deep learning-based perception systems for autonomous driving: A comprehensive survey,” Neurocomput- ing, vol. 489, pp. 255–270, 2022

    work page 2022

  4. [4]

    Tesla’s autopilot was engaged when model 3 crashed into truck, report states,

    “Tesla’s autopilot was engaged when model 3 crashed into truck, report states,” 2019. [Online]. Available: https://www.theverge.com/2019/5/16/18627766/tesla-autopil ot-fatal-crash-delray-florida-ntsb-model-3

    work page 2019

  5. [5]

    Safety of perception systems for automated driving: A case study on apollo,

    S. Kochanthara, T. Singh, A. Forrai, and L. Cleophas, “Safety of perception systems for automated driving: A case study on apollo,” ACM T rans. Softw. Eng. Methodol., vol. 33, no. 3, pp. 64:1– 64:28, 2024. [Online]. Available: https://doi.org/10.1145/3631969

    work page doi:10.1145/3631969 2024

  6. [6]

    Cooperative perception technology of autonomous driving in the internet of vehicles environment: A review,

    G. Cui, W. Zhang, Y. Xiao, L. Yao, and Z. Fang, “Cooperative perception technology of autonomous driving in the internet of vehicles environment: A review,” Sensors, vol. 22, no. 15, p. 5535, 2022

    work page 2022

  7. [7]

    Macp: Efficient model adaptation for cooperative perception,

    Y. Ma, J. Lu, C. Cui, S. Zhao, X. Cao, W. Ye, and Z. Wang, “Macp: Efficient model adaptation for cooperative perception,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2024, pp. 3373–3382. 12

    work page 2024

  8. [8]

    Collaborative perception for connected and autonomous driving: Challenges, possible solutions and opportunities,

    S. Hu, Z. Fang, Y. Deng, X. Chen, and Y. Fang, “Collaborative perception for connected and autonomous driving: Challenges, possible solutions and opportunities,” CoRR, vol. abs/2401.01544,

    arXiv

Show all 52 references
  1. [9]

    Astri and baidu apollo collaborate to promote the implementation of c-v2x technology in hong kong,

    “Astri and baidu apollo collaborate to promote the implementation of c-v2x technology in hong kong,” https://ww w.astri.org/news-detail/astri-and-baidu-apollo-collaborate-to-p romote-the-implementation-of-c-v2x-technology-in-hong-kong/, 2023

    2023

  2. [10]

    Available: https://doi.org/10.48550/arXiv.2401

    [Online]. Available: https://doi.org/10.48550/arXiv.2401. 01544

    doi

  3. [11]

    Collaborative perception in autonomous driving: Methods, datasets, and challenges,

    Y. Han, H. Zhang, H. Li, Y. Jin, C. Lang, and Y. Li, “Collaborative perception in autonomous driving: Methods, datasets, and challenges,” IEEE Intell. T ransp. Syst. Mag. , vol. 15, no. 6, pp. 131–151, 2023. [Online]. Available: https://doi.org/10.1109/MITS.2023.3298534

    2023

  4. [12]

    V2v and v2x technology paves the way for autonomous driv- ing,

    “V2v and v2x technology paves the way for autonomous driv- ing,” https://www.assemblymag.com/articles/98013-v2v-and-v 2x-technology-paves-the-way-for-autonomous-driving, 2023

    2023

  5. [13]

    V2v4real: A real-world large-scale dataset for vehicle-to-vehicle cooperative perception,

    R. Xu, X. Xia, J. Li, H. Li, S. Zhang, Z. Tu, Z. Meng, H. Xiang, X. Dong, R. Song et al. , “V2v4real: A real-world large-scale dataset for vehicle-to-vehicle cooperative perception,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, p...

    2023

  6. [14]

    V2X cooperative perception for autonomous driving: Recent advances and challenges,

    T. Huang, J. Liu, X. Zhou, D. C. Nguyen, M. R. Azghadi, Y. Xia, Q. Han, and S. Sun, “V2X cooperative perception for autonomous driving: Recent advances and challenges,” CoRR, vol. abs/2310.03525, 2023. [Online]. Available: https: //doi.org/10.48550/arXiv.2310.03525

    2023 doi arXiv

  7. [15]

    Cootest: An automated testing approach for V2X communication systems,

    A. Guo, X. Gao, Z. Chen, Y. Xiao, J. Liu, X. Ge, W. Sun, and C. Fang, “Cootest: An automated testing approach for V2X communication systems,” in Proceedings of the 33rd ACM SIGSOFT International Symposium on Software T esting and Analysis, ISSTA 2024, Vienna, Austria, Septembe...

    2024

  8. [16]

    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 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, ...

    2022

  9. [17]

    Lirtest: augmenting lidar point clouds for automated testing of autonomous driving systems,

    A. Guo, Y. Feng, and Z. Chen, “Lirtest: augmenting lidar point clouds for automated testing of autonomous driving systems,” in ISSTA ’22: 31st ACM SIGSOFT International Symposium on Software T esting and Analysis, Virtual Event, South Korea, July 18 - 22, 2022 , S. Ryu and Y. ...

    2022

  10. [18]

    Taulim: Test data augmentation of lidar point cloud by metamorphic relation,

    J. Lin, J. Liu, Q. Zhang, X. Zhang, and C. Fang, “Taulim: Test data augmentation of lidar point cloud by metamorphic relation,” in 44th IEEE/ACM International Conference on Software Engineering: Companion Proceedings, ICSE Companion 2022, Pittsburgh, P A, USA, May 22-24, 2022 ...

    2022

  11. [19]

    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) . IEEE, 2022, pp. 2583–2589

    2022

  12. [20]

    Generating realistic and diverse tests for lidar-based perception systems,

    G. Christian, T. Woodlief, and S. Elbaum, “Generating realistic and diverse tests for lidar-based perception systems,” in 2023 IEEE/ACM 45th International Conference on Software Engineering (ICSE). IEEE, 2023, pp. 2604–2616. [Online]. Available: https://doi.org/10.1109/ICSE486...

    2023

  13. [21]

    V2x-communication- aided autonomous driving: System design and experimental vali- dation,

    C. Jung, D. Lee, S. Lee, and D. H. Shim, “V2x-communication- aided autonomous driving: System design and experimental vali- dation,” Sensors, vol. 20, no. 10, p. 2903, 2020

    2020

  14. [22]

    Enhancements of v2x communication in support of cooperative autonomous driving,

    L. Hobert, A. Festag, I. Llatser, L. Altomare, F. Visintainer, and A. Kovacs, “Enhancements of v2x communication in support of cooperative autonomous driving,” IEEE communications magazine , vol. 53, no. 12, pp. 64–70, 2015

    2015

  15. [23]

    Collaborative automated driv- ing: A machine learning-based method to enhance the accuracy of shared information,

    Z. Y. Rawashdeh and Z. Wang, “Collaborative automated driv- ing: A machine learning-based method to enhance the accuracy of shared information,” in 2018 21st International Conference on Intelligent T ransportation Systems (ITSC) . IEEE, 2018, pp. 3961– 3966

    2018

  16. [24]

    Cooper: Cooperative perception for connected autonomous vehicles based on 3d point clouds,

    Q. Chen, S. Tang, Q. Yang, and S. Fu, “Cooper: Cooperative perception for connected autonomous vehicles based on 3d point clouds,” in Proceedings of the 39th International Conference on Dis- tributed Computing Systems. Dallas, TX, USA: IEEE, July 7-10 2019, pp. 514–524

    2019

  17. [25]

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

    T.-H. Wang, S. Manivasagam, M. Liang, B. Yang, W. Zeng, and R. Urtasun, “V2vnet: Vehicle-to-vehicle communication for joint perception and prediction,” in Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part II 16 . Springer, ...

    2020

  18. [26]

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

    Q. Chen, X. Ma, S. Tang, J. Guo, Q. Yang, and S. Fu, “F-cooper: feature based cooperative perception for autonomous vehicle edge computing system using 3d point clouds,” in Proceedings of the 4th Symposium on Edge Computing . Arlington, Virginia, USA: ACM, November 7-9 2019, p...

    2019

  19. [27]

    Salsanext: Fast, uncertainty-aware semantic segmentation of lidar point clouds,

    T. Cortinhal, G. Tzelepis, and E. E. Aksoy, “Salsanext: Fast, uncertainty-aware semantic segmentation of lidar point clouds,” in Advances in Visual Computing - 15th International Symposium, ISVC 2020, San Diego, CA, USA, October 5-7, 2020, Proceedings, Part II , ser. Lecture N...

    2020 doi

  20. [28]

    A survey on perfor- mance metrics for object-detection algorithms,

    R. Padilla, S. L. Netto, and E. A. Da Silva, “A survey on perfor- mance metrics for object-detection algorithms,” in 2020 interna- tional conference on systems, signals and image processing (IWSSIP) . IEEE, 2020, pp. 237–242

    2020

  21. [29]

    Common corruption robustness of point cloud detectors: Benchmark and enhancement,

    S. Li, Z. Wang, F. Juefei-Xu, Q. Guo, X. Li, and L. Ma, “Common corruption robustness of point cloud detectors: Benchmark and enhancement,” IEEE T ransactions on Multimedia, 2023

    2023

  22. [30]

    Train in germany, test in the USA: making 3d object detectors generalize,

    Y. Wang, X. Chen, Y. You, L. E. Li, B. Hariharan, M. E. Campbell, K. Q. Weinberger, and W. Chao, “Train in germany, test in the USA: making 3d object detectors generalize,” in 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2020, Seattle, WA, USA, Jun...

    2020

  23. [31]

    Metamorphic testing: A review of challenges and opportunities,

    T. Y. Chen, F.-C. Kuo, H. Liu, P .-L. Poon, D. Towey, T. Tse, and Z. Q. Zhou, “Metamorphic testing: A review of challenges and opportunities,” ACM Computing Surveys (CSUR) , vol. 51, no. 1, pp. 1–27, 2018

    2018

  24. [32]

    Cooperative perception and localization for cooperative driving,

    A. Miller, K. Rim, P . Chopra, P . Kelkar, and M. Likhachev, “Cooperative perception and localization for cooperative driving,” in 2020 IEEE International Conference on Robotics and Automation, ICRA 2020, Paris, France, May 31 - August 31, 2020 . IEEE, 2020, pp. 1256–1262. [On...

    2020

  25. [33]

    Shapenet: An information-rich 3d model repository,

    A. X. Chang, T. Funkhouser, L. Guibas, P . Hanrahan, Q. Huang, Z. Li, S. Savarese, M. Savva, S. Song, H. Su et al. , “Shapenet: An information-rich 3d model repository,” arXiv preprint arXiv:1512.03012, 2015

    2015 arXiv

  26. [34]

    Metamorphic testing: A new approach for generating next test cases,

    T. Y. Chen, S. C. Cheung, and S. Yiu, “Metamorphic testing: A new approach for generating next test cases,” CoRR, vol. abs/2002.12543, no. 1, pp. 1–11, 2020

    2002 arXiv

  27. [35]

    Supplementary website,

    “Supplementary website,” https://sites.google.com/view/v2xge n/, 2025

    2025

  28. [36]

    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 European conference on computer vision . Springer, 2022, pp. 107–124

    2022

  29. [37]

    Learning to generate realistic lidar point clouds,

    V . Zyrianov, X. Zhu, and S. Wang, “Learning to generate realistic lidar point clouds,” in Computer Vision - ECCV 2022 - 17th European Conference, T el Aviv, Israel, October 23-27, 2022, Proceedings, Part XXIII , ser. Lecture Notes in Computer Science, S. Avidan, G. J. Brostow...

    2022 doi

  30. [38]

    The pascal visual object classes (voc) challenge,

    M. Everingham, L. Van Gool, C. K. Williams, J. Winn, and A. Zisserman, “The pascal visual object classes (voc) challenge,” International journal of computer vision , vol. 88, pp. 303–338, 2010. [Online]. Available: https://doi.org/10.1007/s11263-009-0275-4

    2010 doi

  31. [39]

    A wilcoxon-type test for trend,

    J. Cuzick, “A wilcoxon-type test for trend,” Statistics in medicine , vol. 4, no. 1, pp. 87–90, 1985

    1985

  32. [40]

    Rangenet ++: Fast and accurate lidar semantic segmentation,

    A. Milioto, I. Vizzo, J. Behley, and C. Stachniss, “Rangenet ++: Fast and accurate lidar semantic segmentation,” in 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2019, Macau, SAR, China, November 3- 13 8, 2019 . IEEE, 2019, pp. 4213–4220. [Onli...

    2019

  33. [41]

    Can offline testing of deep neural networks replace their online testing?

    F. U. Haq, D. Shin, S. Nejati, and L. C. Briand, “Can offline testing of deep neural networks replace their online testing?” Empir. Softw. Eng. , vol. 26, no. 5, p. 90, 2021. [Online]. Available: https://doi.org/10.1007/s10664-021-09982-4

    2021 doi

  34. [42]

    Model vs system level testing of autonomous driving systems: a replication and extension study,

    A. Stocco, B. Pulfer, and P . Tonella, “Model vs system level testing of autonomous driving systems: a replication and extension study,” Empir. Softw. Eng. , vol. 28, no. 3, p. 73, 2023. [Online]. Available: https://doi.org/10.1007/s10664-023-10306-x

    2023 doi

  35. [43]

    Towards the robustness of multiple object tracking systems,

    X. Xie, Y. Duan, S. Chen, and J. Xuan, “Towards the robustness of multiple object tracking systems,” in 2022 IEEE 33rd International Symposium on Software Reliability Engineering (ISSRE). IEEE, 2022, pp. 402–413. [Online]. Available: https://doi.org/10.1109/ISSRE55969.2022.00046

    2022

  36. [44]

    Metamorphic object insertion for testing object detection systems,

    S. Wang and Z. Su, “Metamorphic object insertion for testing object detection systems,” in 35th IEEE/ACM International Conference on Automated Software Engineering, ASE 2020, Melbourne, Australia, September 21-25, 2020 . IEEE, 2020, pp. 1053–1065. [Online]. Available: https://...

    2020

  37. [45]

    Multitest: Physical-aware object insertion for testing multi-sensor fusion perception systems,

    ——, “Multitest: Physical-aware object insertion for testing multi-sensor fusion perception systems,” in Proceedings of the 46th IEEE/ACM International Conference on Software Engineering, ICSE 2024, Lisbon, Portugal, April 14-20, 2024 . ACM, 2024, pp. 139:1–139:13. [Online]. Av...

    2024

  38. [46]

    Benchmarking robustness of ai-enabled multi-sensor fusion systems: Challenges and opportunities,

    X. Gao, Z. Wang, Y. Feng, L. Ma, Z. Chen, and B. Xu, “Benchmarking robustness of ai-enabled multi-sensor fusion systems: Challenges and opportunities,” in Proceedings of the 31st ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software En...

    2023

  39. [47]

    Testing and validating machine learning classifiers by metamorphic testing,

    X. Xie, J. W. K. Ho, C. Murphy, G. E. Kaiser, B. Xu, and T. Y. Chen, “Testing and validating machine learning classifiers by metamorphic testing,” J. Syst. Softw. , vol. 84, no. 4, pp. 544–558,

  40. [48]

    A survey on metamorphic testing,

    S. Segura, G. Fraser, A. B. Sanchez, and A. Ruiz-Cort ´es, “A survey on metamorphic testing,” IEEE T ransactions on software engineering, vol. 42, no. 9, pp. 805–824, 2016. [Online]. Available: https://doi.org/10.1109/TSE.2016.2532875

    2016

  41. [49]

    Semantic-guided fuzzing for virtual testing of autonomous driving systems,

    A. Guo, Y. Feng, Y. Cheng, and Z. Chen, “Semantic-guided fuzzing for virtual testing of autonomous driving systems,” J. Syst. Softw. , vol. 212, p. 112017, 2024. [Online]. Available: https://doi.org/10.1016/j.jss.2024.112017

    2024

  42. [51]

    Metamorphic testing of driverless cars,

    Z. Q. Zhou and L. Sun, “Metamorphic testing of driverless cars,” Commun. ACM, vol. 62, no. 3, pp. 61–67, 2019. [Online]. Available: https://doi.org/10.1145/3241979

    2019 doi

  43. [2011]

    Available: https://doi.org/10.1016/j.jss.2010.11

    [Online]. Available: https://doi.org/10.1016/j.jss.2010.11. 920

    2010 doi

  44. [2024]

    Available: https://doi.org/10.1016/j.eswa.2023

    [Online]. Available: https://doi.org/10.1016/j.eswa.2023. 122836

    2023 doi

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.