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

We study the effect of adversarial perturbations on the task of monoculardepth prediction. Specifically, we explore the ability of small, imperceptibleadditive perturbations to selectively alter the perceived geometry of thescene. We show that such perturbations can not only globally re-scale thepredicted distances from the camera, but also alter the prediction to match adifferent target scene. We also show that, when given semantic or instanceinformation, perturbations can fool the network to alter the depth of specificcategories or instances in the scene, and even remove them while preserving therest of the scene. To understand the effect of targeted perturbations, weconduct experiments on state-of-the-art monocular depth prediction methods. Ourexperiments reveal vulnerabilities in monocular depth prediction networks, andshed light on the biases and context learned by them.

Targeted Adversarial Perturbations for Monocular Depth Prediction