Pith. sign in

REVIEW 2 cited by

Uncertainty Estimation for 3D Object Detection via Evidential Learning

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 2410.23910 v1 pith:IYWBLU53 submitted 2024-10-31 cs.CV

Uncertainty Estimation for 3D Object Detection via Evidential Learning

classification cs.CV
keywords uncertaintydetectionobjectestimatesscenesdetectorevidentialframework
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

3D object detection is an essential task for computer vision applications in autonomous vehicles and robotics. However, models often struggle to quantify detection reliability, leading to poor performance on unfamiliar scenes. We introduce a framework for quantifying uncertainty in 3D object detection by leveraging an evidential learning loss on Bird's Eye View representations in the 3D detector. These uncertainty estimates require minimal computational overhead and are generalizable across different architectures. We demonstrate both the efficacy and importance of these uncertainty estimates on identifying out-of-distribution scenes, poorly localized objects, and missing (false negative) detections; our framework consistently improves over baselines by 10-20% on average. Finally, we integrate this suite of tasks into a system where a 3D object detector auto-labels driving scenes and our uncertainty estimates verify label correctness before the labels are used to train a second model. Here, our uncertainty-driven verification results in a 1% improvement in mAP and a 1-2% improvement in NDS.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

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

  1. Uncertainty-Aware Pedestrian Attribute Recognition via Evidential Deep Learning

    cs.CV 2026-04 unverdicted novelty 6.0

    UAPAR is the first evidential deep learning framework for pedestrian attribute recognition that estimates attribute-wise epistemic uncertainty via a region-aware module and uses uncertainty-guided curriculum learning ...

  2. UECP: Uncertainty-Enhanced Collaborative Perception

    cs.CV 2026-06 unverdicted novelty 5.0

    UECP replaces detection-correlated confidence maps with a LiDAR point-density uncertainty map and introduces Uncertainty-Aware Pyramid Fusion to improve collaborative perception.