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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

When deploying modern machine learning-enabled robotic systems in high-stakesapplications, detecting distribution shift is critical. However, most existingmethods for detecting distribution shift are not well-suited to roboticssettings, where data often arrives in a streaming fashion and may be veryhigh-dimensional. In this work, we present an online method for detectingdistribution shift with guarantees on the false positive rate - i.e., whenthere is no distribution shift, our system is very unlikely (with probability$< \epsilon$) to falsely issue an alert; any alerts that are issued shouldtherefore be heeded. Our method is specifically designed for efficientdetection even with high dimensional data, and it empirically achieves up to11x faster detection on realistic robotics settings compared to prior workwhile maintaining a low false negative rate in practice (whenever there is adistribution shift in our experiments, our method indeed emits an alert). Wedemonstrate our approach in both simulation and hardware for a visual servoingtask, and show that our method indeed issues an alert before a failure occurs.

Online Distribution Shift Detection via Recency Prediction