English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

The robustness of 3D perception systems under natural corruptions fromenvironments and sensors is pivotal for safety-critical applications. Existinglarge-scale 3D perception datasets often contain data that are meticulouslycleaned. Such configurations, however, cannot reflect the reliability ofperception models during the deployment stage. In this work, we present Robo3D,the first comprehensive benchmark heading toward probing the robustness of 3Ddetectors and segmentors under out-of-distribution scenarios against naturalcorruptions that occur in real-world environments. Specifically, we considereight corruption types stemming from severe weather conditions, externaldisturbances, and internal sensor failure. We uncover that, although promisingresults have been progressively achieved on standard benchmarks,state-of-the-art 3D perception models are at risk of being vulnerable tocorruptions. We draw key observations on the use of data representations,augmentation schemes, and training strategies, that could severely affect themodel's performance. To pursue better robustness, we propose adensity-insensitive training framework along with a simple flexiblevoxelization strategy to enhance the model resiliency. We hope our benchmarkand approach could inspire future research in designing more robust andreliable 3D perception models. Our robustness benchmark suite is publiclyavailable.

Robo3D: Towards Robust and Reliable 3D Perception against Corruptions