Pith. sign in

REVIEW 5 cited by

RoboBEV: Towards Robust Bird's Eye View Perception under Corruptions

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2304.06719 v1 pith:6REVPRPF submitted 2023-04-13 cs.CV cs.RO

classification cs.CVcs.RO
keywords robustnessacrossbirdcorruptionsfindingsmodelsout-of-distributionperception
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

The recent advances in camera-based bird's eye view (BEV) representation exhibit great potential for in-vehicle 3D perception. Despite the substantial progress achieved on standard benchmarks, the robustness of BEV algorithms has not been thoroughly examined, which is critical for safe operations. To bridge this gap, we introduce RoboBEV, a comprehensive benchmark suite that encompasses eight distinct corruptions, including Bright, Dark, Fog, Snow, Motion Blur, Color Quant, Camera Crash, and Frame Lost. Based on it, we undertake extensive evaluations across a wide range of BEV-based models to understand their resilience and reliability. Our findings indicate a strong correlation between absolute performance on in-distribution and out-of-distribution datasets. Nonetheless, there are considerable variations in relative performance across different approaches. Our experiments further demonstrate that pre-training and depth-free BEV transformation has the potential to enhance out-of-distribution robustness. Additionally, utilizing long and rich temporal information largely helps with robustness. Our findings provide valuable insights for designing future BEV models that can achieve both accuracy and robustness in real-world deployments.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 5 Pith papers

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

  1. Iterate or Widen? When Test-Time Refinement Helps LiDAR Scene Completion: A Controlled Study of Evidence Geometry, Training Coverage, and Compute

    cs.CV 2026-08 conditional novelty 6.0 of 10

    A controlled LiDAR scene completion study finds that fixed-depth iterative refinement beats a parameter-matched wide model only for coherent missing regions, while training augmentation dominates for scattered thinning.

  2. SafeMap: Robust HD Map Construction from Incomplete Observations

    cs.CV 2025-07 conditional novelty 6.0 of 10

    SafeMap improves HD map construction accuracy under missing camera views by reconstructing the missing perspective features with Gaussian-sampled attention and correcting the BEV features through distillation.

  3. SAM4D: Segment Anything in Camera and LiDAR Streams

    cs.CV 2025-06 conditional novelty 6.0 of 10

    SAM4D is a promptable model that segments and tracks objects across camera and LiDAR streams with cross-modal prompts, trained on pseudo-labels generated by an automated data engine.

  4. Foundation Models for Autonomous Driving Perception: A Survey Through Core Capabilities

    cs.RO 2025-09 conditional novelty 4.0 of 10

    Foundation-model perception for autonomous driving is surveyed through four capability lenses: generalized knowledge, spatial understanding, multi-sensor robustness, and temporal understanding.

  5. What Really Matters for Robust Multi-Sensor HD Map Construction?

    cs.CV 2025-07 conditional novelty 4.0 of 10

    Combining data augmentation, cross-modal attention fusion, and modality dropout training improves robustness of camera-LiDAR HD map construction under 13 synthetic sensor corruptions and raises clean nuScenes mAP to 77.0.

Pith tools