Effort-Based Criticality Metrics for Evaluating 3D Perception Errors in Autonomous Driving

Abhinav Valada; Mario Berk; Sharang Kaul; Simon Bultmann

REVIEW 3 major objections 4 minor 28 references

Effort-Based Criticality Metrics for Evaluating 3D Perception Errors in Autonomous Driving

T0 review · 3 major / 4 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read Perception errors become braking and steering effort, not just counts

desk verdict Useful new metrics, but the headline non-redundancy claim is inflated by an over-permissive collision filter. read the letter →

arxiv 2603.28029 v2 pith:JS7UE2XJ submitted 2026-03-30 cs.CV cs.RO

Sharang Kaul , Simon Bultmann , Mario Berk , Abhinav Valada This is my paper · ORCID

classification cs.CVcs.RO

keywords effort-basedcriticalityfalsespeedreductionmaximumdecelerationratelateralevasionaccelerationreachabilityanalysis3Dperceptionevaluationautonomousdrivingsafetypositivesandnegatives

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that the safety relevance of false-positive and false-negative perception errors in autonomous driving can be quantified as physical collision-avoidance effort: false positives are scored by the cumulative speed reduction they would force (FSR), and false negatives by the peak braking demand they would require (MDR). A third metric, lateral evasion acceleration (LEA), covers cases where steering is the cheaper avoidance channel. The authors argue that these effort-based metrics capture safety-relevant information that conventional time-based or deceleration-based criticality measures like TTC and DRAC miss, and that they can rank and mine the most dangerous perception failures without needing a full planning stack. If true, this would give developers a way to compare detection and tracking pipelines on safety impact rather than raw accuracy.

What carries the argument

The central mechanism is a pair of longitudinal kinematic equations: Eq. (2) computes the instantaneous deceleration caused by a persistent false positive, and Eq. (5) computes the braking demand for a missed object with its own acceleration. FSR aggregates the per-frame braking over the error's duration; MDR takes the per-frame maximum. These are complemented by a reachability-based ellipsoidal collision filter that grows vehicle shapes quadratically under bounded acceleration and predicts collision time as the earliest reachable-set overlap, ensuring that only dynamically plausible threats are scored. LEA, adapted from lateral-evasion kinematics, computes the minimum lateral acceleration n

What would settle it

Run a closed-loop simulation or real-vehicle test where a planner responds to the same false positive and false negative scenarios; compare the actual braking or steering effort to the FSR/MDR/LEA predictions. If the predicted effort does not correlate with the planner's actual commanded deceleration or lateral acceleration across a set of diverse traffic scenes, the metric's claim to represent avoidance effort collapses. A simpler check: vary treact from 0.2 to 0.5 s and amax bounds by ±50% and observe whether the ranking of three pipelines by cumulative FSR/MDR changes materially—if it flips

Watch

Extended reading notes

Core claim

The paper's central claim is that each perception error can be translated, via kinematic equations, into the avoidance effort it would impose on the ego vehicle. For false positives, the required deceleration is derived by equating the distance the ego vehicle travels to the phantom object's distance plus separation until speeds match, yielding a per-frame braking value that accumulates over the phantom's lifetime into FSR. For false negatives, a constant-acceleration model accounts for the missed object's own motion, and the maximum per-frame braking demand becomes MDR. Both are gated by a reachability-based collision filter that predicts whether the ego and object reachable sets overlap wi

Load-bearing premise

The load-bearing premise is that a single longitudinal, constant-deceleration model with fixed scalar parameters (reaction time, acceleration bounds, reachable-set growth) is a faithful proxy for what a real planner would do in response to a perception error.

Editorial extensions

If this is right

Raw detection error counts substantially overestimate safety risk: the paper finds that 65–93% of errors are non-critical, so effort-based metrics could serve as a risk filter before expensive closed-loop testing.
False positives that persist for many frames, even with moderate per-frame braking, accumulate high FSR; this penalizes phantom tracks that no single-frame TTC-based metric would flag.
False-negative severity is nearly independent of the perception pipeline: mean MDR stays at 2.0–2.5 m/s² across all pipelines and datasets, suggesting a geometry-driven safety floor that detector improvements alone cannot remove.
LEA is almost uncorrelated with longitudinal metrics, meaning lateral evasion effort is a distinct risk axis; combining it with MDR can distinguish an in-lane missed vehicle (high MDR, low LEA) from a laterally distant one (elevated MDR, low LEA).
Cumulative effort—the sum of FSR or MDR over all error tracks—can disambiguate pipelines that have similar precision or recall but very different total safety burden.

Reading between the lines

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

The reachability gate admits 98% of FN and 93% of FP tracks, while a tighter SAT gate admits only 4% and 1%; this suggests the reported non-critical fractions and correlation structure depend heavily on choosing a conservative filter, and a different gate could change which errors are considered critical.
Because the metrics use fixed scalar parameters (reaction time 0.3 s, acceleration bounds from naturalistic driving), the rank ordering of pipelines might shift if those parameters were varied; a sensitivity analysis over these values would test the robustness of the cross-pipeline conclusions.
The metrics are derived for longitudinal and lateral evasion separately; combining them into a single avoidance cost (as the paper mentions via a steer-threat number) could produce a more actionable safety score, though the paper intentionally keeps them separate to avoid discounting converging lateral threats.
If the metrics were applied at multiple confidence thresholds in a precision-recall sweep, they could yield criticality-weighted detection costs, effectively turning the framework into a safety-aware benchmark that does not require closed-loop simulation.

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

The core idea here is good: FSR and MDR give interpretable physical meaning to false positives and false negatives, and the evaluation is broad—three pipelines, two datasets, and an honest discussion of the RSB/SAT trade-off. The kinematic derivations in Eqs. (2) and (5) are elementary but sound, and FSR's persistence-sensitivity is a genuinely useful complement to single-frame measures. I'd want to see this paper engaged with seriously.

The main problem is the central empirical claim. The Spearman correlations in Tab. V are computed under RSB, which admits 98% of FN tracks. The paper itself shows (Fig. 4) that this gate scores opposite-road vehicles and other laterally distant objects that a planner would ignore. Under SAT—where the threat set is actually plausible—MDR–DRAC correlation jumps from 0.30 to 0.68. That doesn't tell me MDR is capturing unique safety information; it tells me the low RSB correlations are largely an artifact of scoring a pile of non-threats. The 42% of high-MDR FNs with TET=0 needs to be recomputed on the SAT-gated or lane-filtered set before it can support the non-redundancy claim.

Also, the 65–93% non-critical fractions depend entirely on hand-set thresholds in Tab. III, with no sensitivity analysis. And the "braking impulse" phrasing for FSR is a misnomer—it's a cumulative speed reduction, not an impulse.

That said, the paper is careful in its limitations discussion, acknowledges the metrics are heuristic, and doesn't oversell them as replacements for closed-loop testing. The FSR/MDR definitions themselves are clean and reproducible enough to be worth testing on other datasets.

Verdict: worth sending to peer review, but the authors should be pushed to separate the filter's effect from the metrics' properties—report both gates throughout, or restrict the main correlation claims to the SAT subset. If the complementarity survives that, it's a real contribution. If not, the paper still stands as a practical screening tool, just with a more modest conclusion.

Referee Report

3 major / 4 minor

Summary. The paper proposes a family of effort-based criticality metrics for open-loop evaluation of 3D perception errors in autonomous driving. False Speed Reduction (FSR) translates persistent false positives into a cumulative braking-impulse-like quantity, Maximum Deceleration Rate (MDR) translates missed objects into a peak required longitudinal deceleration, and Lateral Evasion Acceleration (LEA) quantifies the minimum lateral acceleration to steer around a predicted collision. The metrics are gated by an ellipsoidal reachability filter that decides which object–ego pairs are scored. The framework is evaluated on nuScenes and Argoverse 2 with three perception pipelines (AB3DMOT, CenterPoint, BEVFusion), reporting severity distributions, cumulative effort, and Spearman correlations against established metrics such as TTC, DRAC, THW, and TET. The central claims are that 65–93% of perception errors are non-critical and that the proposed metrics capture safety-relevant information not provided by conventional measures.

Significance. If the central claim is substantiated, the paper offers a practically useful way to rank perception failures by the collision-avoidance effort they would impose, with clearly interpretable units (m/s and m/s^2) and explicit error-type separation. The work is strengthened by self-contained kinematic derivations (Eqs. 2 and 5), a modular collision filter that can be exchanged, a broad evaluation across datasets and pipelines, and scenario-level analyses that illustrate how the metrics behave. The paper also honestly lists several limitations, including the heuristic nature of the metrics and the conservatism of the reachability gate. However, the central claim of non-redundancy with established metrics rests on correlation evidence that may be confounded by the very conservatism the authors acknowledge. The paper therefore requires additional validation of the planner-relevance of the scored threat set before the strong abstract claim can be accepted.

major comments (3)

[§IV-D, Table V, and §V (Role of the collision filter)] The complementarity claim is built on Spearman correlations computed under the RSB gate, which admits 98% of FN and 93% of FP tracks. As the paper shows in Fig. 4, many of these are opposite-road or off-road objects that a planner would never brake for. The low MDR–DRAC correlation (ρ=0.30) and the statistic that 42% of MDR>3 m/s² FNs have TET=0 may therefore be dominated by non-threat tracks, not by safety-relevant information. The authors acknowledge the gate's conservatism but still present the resulting correlation structure as evidence of non-redundancy. Please report the correlation analysis on a planner-relevant subset (e.g., ego-lane objects, or the SAT-gated set) and show that the metrics retain their complementary value there. This is load-bearing for the abstract claim.
[§III-B, Eq. (4)] FSR is defined as (N_frames × T_cycle) × a_avg, i.e., the sum of per-frame required decelerations multiplied by the frame interval. The per-frame deceleration is the constant braking that would be needed to avoid the phantom if the ego started braking at that frame. Summing these values over a persistent phantom is not the accumulated velocity loss of a single braking maneuver; it can double-count the same threat across frames and can exceed the physical speed differential (e.g., Fig. 4 FP1: 17.5 m/s). If FSR is intended as a heuristic 'effort' score, the paper should state this more precisely and justify why summation rather than, say, the maximum required speed reduction is the right aggregation. As written, the physical interpretation of 'cumulative velocity loss' is not fully supported.
[Abstract and §VI] The abstract states that Spearman correlation analysis 'confirms that all three metrics capture safety-relevant information inaccessible to established time-based, deceleration-based, or normalized criticality measures.' Given that the paper itself describes the metrics as 'heuristic approximations' (§V) and that no closed-loop planner validation is performed, 'confirms' overstates the evidence. A correlation analysis at track level shows statistical non-redundancy, but it does not establish safety relevance. Please temper the wording or add a validation step (e.g., replaying high-scoring errors in a planner or comparing against human-judged criticality) to support the safety-relevance claim.

minor comments (4)

[§II and References] Reference [24] is listed as 'Argoverse 2' but the text says 'Argoverse [24]' at one point in Section IV-A; please make the citation consistent.
[Table II and Fig. 3] The table and figure captions would benefit from explicitly stating the dataset split (validation set) and the number of scenes used. The current text mentions nuScenes has 1000 scenes, but the exact split is not stated.
[§III-D, Eq. (5)] The reachable-set semi-axis growth is written as σ(τ)=d/2 + 0.5*a_max*τ², but the justification for the linear-in-τ term (from the initial position uncertainty) is not given. Please add a short explanation or reference.
[§IV-D] The text says 'All pairwise |ρ| among established metrics exceed 0.69 (RSB)' but does not list the TET–THW correlation explicitly. A supplementary table with all pairwise correlations would be helpful for reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: FSR/MDR are explicit kinematic constructs with externally sourced parameters; the correlation claims are post-hoc empirical comparisons, not fitted inputs.

full rationale

The central derivation chain is self-contained in the kinematic sense. FSR and MDR are obtained by solving physically stated one-dimensional longitudinal equations (Eqs. 1-2 and Eq. 5) with fixed scalar parameters (treact=0.3s, acceleration bounds in Table I) that are adopted from external sources, not fitted to the evaluation data or to the claimed correlation results. LEA is explicitly adapted from the prior lateral-evasion kinematics of [7] and reachability-based collision timing from [8]; these are not the present authors' own works, so the adoption is not a load-bearing self-citation. The Spearman analysis in Table V is a genuine post hoc comparison between the proposed metrics and established metrics, not a calibration step that forces the reported correlations. The paper candidly discloses the RSB filter's conservatism (Section V: admitting 98% of FN and 93% of FP tracks, versus 4% and 1% for SAT) and the scenario-level cases where laterally distant or opposite-road objects are scored (Section IV-E, Fig. 4). This is a limitation on the safety-relevance interpretation of the empirical results, and a potential confound for the 'inaccessible information' claim, but it is not a circularity: the metrics are not defined in terms of the correlation outcome, and no predicted quantity reduces by construction to a fitted input. The self-references to the authors' own prior work ([3]-[5]) are contextual and do not support any load-bearing uniqueness or derivation claim. Accordingly, no circular step can be identified, and the appropriate score is 0.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new physical entities, forces, or objects. Its free parameters are all configurable constants (reaction time, acceleration bounds, caps, thresholds) and modeling choices. The most consequential free choices are the reachability bounds and the severity thresholds, which directly determine the reported 65–93% non-critical fractions and the correlation structure.

free parameters (5)

Reaction time treact (0.3 s) = 0.3 s
Chosen from prior AV latency literature [23]; directly enters Eqs. (2) and (5) and thus every FSR/MDR value.
Reachable-set acceleration bounds (alon in [-3.0, 2.0], alat,max 2.0 m/s^2) = alon_min=-3.0, alon_max=2.0, alat_max=2.0 m/s^2
Controls the growth of the ellipsoidal reachable sets and the collision gate; a larger bound would flag many more interactions. Adopted from [8] and naturalistic driving data, but still a modeling choice.
Capability caps (abrake,max=10.0, alat,cap=5.0 m/s^2) = 10.0, 5.0 m/s^2
Caps for FSR/MDR/LEA; chosen to reflect emergency braking and rollover limits from [21], [22]. Affects worst-case and cumulative values (many entries saturate at 10.0).
Horizon Thorizon=5.0 s, dt=0.1 s = 5.0 s / 0.1 s
Collision detection horizon; a longer horizon would flag more tracks.
Severity thresholds (MDR/FSR/LEA zones in Table III) = e.g., MDR 2/4/6 m/s^2; FSR 1/2.5/5 m/s; LEA 1/2/4 m/s^2
Chosen by hand and from driving guidelines; the headline '65–93% non-critical' numbers depend on these thresholds, with no sensitivity analysis.

assumptions (6)

domain assumption Longitudinal kinematics: the minimum required deceleration is computed by matching travel distances under constant deceleration and a reaction-time delay (Eqs. 1-5).
Assumes the ego reacts with a fixed delay and then applies constant deceleration; real planners use more complex trajectories. This is the core of FSR/MDR.
domain assumption Missed object's acceleration a_FN is known/estimated and incorporated in Eq. 5; constant-velocity model for FPs.
For real vehicles this is reasonable (velocity/acceleration from tracking), but for phantom objects the constant-velocity assumption is an assumption about ghost detections.
domain assumption Reachable sets of ego and object are ellipsoidal over-approximations with quadratic growth, and collision occurs when they overlap.
The ellipsoidal model is a known over-approximation; the paper relies on arguments from reachability analysis for safety-related claims but does not prove the over-approximation tightness for its specific setup.
domain assumption Lateral evasion can be modeled as straight-line constant-lateral-acceleration maneuvers (widen/cross strategies) with a fixed clearance wc.
The lane-change geometry and velocities are simplified; real steering has vehicle dynamics constraints beyond constant lateral acceleration.
ad hoc to paper Hungarian matching with 2.0 m center-distance threshold correctly separates FPs from FNs.
The paper admits matching influences criticality, since a 2.1 m off-center object can become a high-MDR FN; the choice affects all downstream numbers.
domain assumption Ground-truth annotations and track identities from nuScenes/Argoverse 2 are correct.
Standard dataset assumption; errors in ground truth would propagate into every metric.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Effort-Based Criticality Metrics for Evaluating 3D Perception Errors in Autonomous Driving." pith.science (2026). https://pith.science/paper/JS7UE2XJ

@misc{pith2026260328029,
  author       = {Pith},
  title        = {Pith review of: Effort-Based Criticality Metrics for Evaluating 3D Perception Errors in Autonomous Driving},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JS7UE2XJ}},
  note         = {Machine review of arXiv:2603.28029}
}

read the original abstract

Criticality metrics such as time-to-collision (TTC) quantify collision urgency but do not distinguish the operational consequences of false-positive (FP) and false-negative (FN) perception errors. We formulate two error-specific effort metrics: False Speed Reduction (FSR), the cumulative velocity loss associated with persistent phantom detections, and Maximum Deceleration Rate (MDR), the peak braking demand associated with missed objects under a longitudinal kinematic model. These longitudinal metrics are complemented by Lateral Evasion Acceleration (LEA), adapted from prior lateral-evasion kinematics and coupled with reachability-based collision timing. The collision check quantifies the minimum steering effort required to avoid a predicted collision. A dynamically conservative, semantically unfiltered reachability gate selects candidate interactions before frame-level scoring and track-level aggregation. Evaluation on nuScenes and Argoverse 2 shows that 65% to 93% of errors fall below the chosen criticality thresholds. Correlation and threshold analysis indicate that the proposed metrics provide complementary rankings for screening and mining perception failures and are not substitutes for closed-loop safety validation.

Figures

Figures reproduced from arXiv: 2603.28029 by the authors.

**Figure 1.** Effort-based criticality assessment of perception errors. The pipeline identifies false positives (FPs) and false negatives (FNs) by matching tracker [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗

**Figure 2.** Collision avoidance strategies between the ego vehicle (gray) and a [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Distribution of non-safe perception errors (absolute track counts) on [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Scenario-level analysis for scene 26a6b03c (nuScenes, BEVFusion). Each row shows, from left to right: bird’s-eye trajectory evolution with start/end markers, reachability-based TTC (TTCRSB), braking effort (abrake), lateral evasion effort (LEA), and the cumulative crit…

**Figure 5.** Figure 5: Classical TTC (d/vrel) vs. minimum distance for effort-critical carcategory tracks on the nuScenes validation set. Dashed line: TTC= 2 s. A substantial fraction of MDR-critical FN tracks cluster at TTC>3 s or at the dataset cap of 10 s, indicating that classical TTC r…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 2 linked inside Pith

[1]

nuscenes revisited: Progress and challenges in autonomous driving,

W. K. Fong, V . E. Liong, K. S. Tan, and H. Caesar, “nuscenes revisited: Progress and challenges in autonomous driving,”arXiv preprint arXiv:2512.02448, 2025

arXiv 2025
[2]

3d multi-object tracking: A baseline and new evaluation metrics,

X. Weng, J. Wang, D. Held, and K. Kitani, “3d multi-object tracking: A baseline and new evaluation metrics,” inIEEE/RSJ International Conference on Intelligent Robots and Systems, 2020, pp. 10 359–10 366

2020
[3]

Robust object detection using knowledge graph embeddings,

C. Lang, A. Braun, and A. Valada, “Robust object detection using knowledge graph embeddings,” inDAGM German Conference on Pattern Recognition, 2022, pp. 445–461

2022
[4]

Progressive multi- modal fusion for robust 3d object detection,

R. Mohan, D. Cattaneo, F. Drews, and A. Valada, “Progressive multi- modal fusion for robust 3d object detection,” in8th Annual Conference on Robot Learning, 2024

2024
[5]

Efficient learn- ing of urban driving policies using bird’s-eye-view state representations,

R. Trumpp, M. Büchner, A. Valada, and M. Caccamo, “Efficient learn- ing of urban driving policies using bird’s-eye-view state representations,” inIEEE Int. Conf. on Intelligent Transportation Systems, 2023

2023
[6]

Criticality metrics for automated driving: A review and suitability analysis of the state of the art,

L. Westhofen,et al., “Criticality metrics for automated driving: A review and suitability analysis of the state of the art,”Archives of Computational Methods in Engineering, vol. 30, no. 1, p. 1–35, 2022

2022
[7]

Acceleration-based collision criticality metric for holistic online safety assessment in automated driving,

C. Wang, C. Popp, and H. Winner, “Acceleration-based collision criticality metric for holistic online safety assessment in automated driving,”IEEE Access, vol. 10, pp. 70 662–70 674, 2022

2022
[8]

Towards threat metric evaluation in complex urban scenarios,

P. Schneider, M. Butz, C. Heinzemann, J. Oehlerking, and M. Woehrle, “Towards threat metric evaluation in complex urban scenarios,” inIEEE International Intelligent Transportation Systems Conference, 2021

2021

Show all 28 references

[9]

Computing reachable sets of hybrid systems using a combination of zonotopes and polytopes,

M. Althoff, O. Stursberg, and M. Buss, “Computing reachable sets of hybrid systems using a combination of zonotopes and polytopes,” Nonlinear analysis: hybrid systems, vol. 4, no. 2, pp. 233–249, 2010

2010
[10]

Safety-aware metric for people detection,

M. Wolf, L. R. Douat, and M. Erz, “Safety-aware metric for people detection,” inIEEE Int. Conf. Intelligent Transportation Systems, 2021

2021
[11]

A comprehensive safety metric to evaluate perception in autonomous systems,

G. V olket al., “A comprehensive safety metric to evaluate perception in autonomous systems,” inIEEE Int. Conf. on Intelligent Transportation Systems, 2020

2020
[12]

Evaluating object (mis)detection from a safety and reliability perspective: Discussion and measures,

A. Ceccarelli and L. Montecchi, “Evaluating object (mis)detection from a safety and reliability perspective: Discussion and measures,” IEEE Access, vol. 11, p. 44952–44963, 2023

2023
[13]

Learning to evaluate perception models using planner-centric metrics,

J. Philion, A. Kar, and S. Fidler, “Learning to evaluate perception models using planner-centric metrics,” inIEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 14 052–14 061

2020
[14]

Usc: Uncom- promising spatial constraints for safety-oriented 3d object detectors in autonomous driving,

B. H.-C. Liao, C.-H. Cheng, H. Esen, and A. Knoll, “Usc: Uncom- promising spatial constraints for safety-oriented 3d object detectors in autonomous driving,” inIEEE International Conference on Intelligent Transportation Systems, 2024, pp. 3466–3472

2024
[15]

Ec-iou: Orienting safety for object detectors via ego-centric intersection-over-union,

——, “Ec-iou: Orienting safety for object detectors via ego-centric intersection-over-union,” inIEEE/RSJ International Conference on Intelligent Robots and Systems, 2024, pp. 9439–9446

2024
[16]

Towards safety-aware pedestrian detection in autonomous systems,

M. Lyssenko, C. Gladisch, C. Heinzemann, M. Woehrle, and R. Triebel, “Towards safety-aware pedestrian detection in autonomous systems,” in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2022, pp. 293–300

2022
[17]

A flow-based credibility metric for safety-critical pedestrian detection,

——, “A flow-based credibility metric for safety-critical pedestrian detection,” inInternational Conference on Computer Safety, Reliability, and Security, 2024

2024
[18]

A safety-adapted loss for pedestrian detection in autonomous driving,

M. Lyssenko, P. Pimplikar, M. Bieshaar, F. Nozarian, and R. Triebel, “A safety-adapted loss for pedestrian detection in autonomous driving,” IEEE International Conference on Robotics and Automation, 2024

2024
[19]

Safety-aware hardening of 3d object detection neural network systems,

C.-H. Cheng, “Safety-aware hardening of 3d object detection neural network systems,”arXiv preprint arXiv:2003.11242, 2020

2003 arXiv
[20]

Collision detection and optimized path design for dragon dance team based on separated axis theorem (sat),

J. Zhao, “Collision detection and optimized path design for dragon dance team based on separated axis theorem (sat),” inInternational Conference on Big Data, Information and Computer Network, 2025

2025
[21]

Towards standardization of av safety: C++ library for responsibility sensitive safety,

B. Gassmann,et al., “Towards standardization of av safety: C++ library for responsibility sensitive safety,” inIEEE Intelligent V ehicles Symposium, 2019, pp. 2265–2271

2019
[22]

Rollover and near-rollover kinematics during evasive steer maneuvers,

C. Young, D. King, and G. Siegmund, “Rollover and near-rollover kinematics during evasive steer maneuvers,”SAE Int. Journal of Advances and Current Practices in Mobility, 2022

2022
[23]

Plm-net: Perception latency mitigation network for vision-based lateral control of autonomous vehicles,

A. Khalil and J. Kwon, “Plm-net: Perception latency mitigation network for vision-based lateral control of autonomous vehicles,”arXiv preprint arXiv:2407.16740, 2024

2024
[24]

Argoverse 2: Next generation datasets for self-driving perception and forecasting,

B. Wilson,et al., “Argoverse 2: Next generation datasets for self-driving perception and forecasting,”arXiv preprint arXiv:2301.00493, 2023

2023 arXiv
[25]

Center-based 3d object detection and tracking,

T. Yin, X. Zhou, and P. Krähenbühl, “Center-based 3d object detection and tracking,” inIEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 11 779–11 788

2021
[26]

Bevfusion: Multi-task multi-sensor fusion with unified bird’s-eye view representation,

Z. Liu,et al., “Bevfusion: Multi-task multi-sensor fusion with unified bird’s-eye view representation,” inIEEE International Conference on Robotics and Automation, 2023, pp. 2774–2781

2023
[27]

Assessment protocol-safety assist,

N. Euro, “Assessment protocol-safety assist,”Test protocol V ersion, vol. 6, 2015

2015
[28]

Criticality metric for the safety validation of automated driving using model predictive trajectory optimization,

P. Junietzet al., “Criticality metric for the safety validation of automated driving using model predictive trajectory optimization,” inInt. conf. on intelligent transportation systems, 2018

2018

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.

[1] [1]

nuscenes revisited: Progress and challenges in autonomous driving,

W. K. Fong, V . E. Liong, K. S. Tan, and H. Caesar, “nuscenes revisited: Progress and challenges in autonomous driving,”arXiv preprint arXiv:2512.02448, 2025

arXiv 2025

[2] [2]

3d multi-object tracking: A baseline and new evaluation metrics,

X. Weng, J. Wang, D. Held, and K. Kitani, “3d multi-object tracking: A baseline and new evaluation metrics,” inIEEE/RSJ International Conference on Intelligent Robots and Systems, 2020, pp. 10 359–10 366

2020

[3] [3]

Robust object detection using knowledge graph embeddings,

C. Lang, A. Braun, and A. Valada, “Robust object detection using knowledge graph embeddings,” inDAGM German Conference on Pattern Recognition, 2022, pp. 445–461

2022

[4] [4]

Progressive multi- modal fusion for robust 3d object detection,

R. Mohan, D. Cattaneo, F. Drews, and A. Valada, “Progressive multi- modal fusion for robust 3d object detection,” in8th Annual Conference on Robot Learning, 2024

2024

[5] [5]

Efficient learn- ing of urban driving policies using bird’s-eye-view state representations,

R. Trumpp, M. Büchner, A. Valada, and M. Caccamo, “Efficient learn- ing of urban driving policies using bird’s-eye-view state representations,” inIEEE Int. Conf. on Intelligent Transportation Systems, 2023

2023

[6] [6]

Criticality metrics for automated driving: A review and suitability analysis of the state of the art,

L. Westhofen,et al., “Criticality metrics for automated driving: A review and suitability analysis of the state of the art,”Archives of Computational Methods in Engineering, vol. 30, no. 1, p. 1–35, 2022

2022

[7] [7]

Acceleration-based collision criticality metric for holistic online safety assessment in automated driving,

C. Wang, C. Popp, and H. Winner, “Acceleration-based collision criticality metric for holistic online safety assessment in automated driving,”IEEE Access, vol. 10, pp. 70 662–70 674, 2022

2022

[8] [8]

Towards threat metric evaluation in complex urban scenarios,

P. Schneider, M. Butz, C. Heinzemann, J. Oehlerking, and M. Woehrle, “Towards threat metric evaluation in complex urban scenarios,” inIEEE International Intelligent Transportation Systems Conference, 2021

2021