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

Our field has recently witnessed an arms race of neural network-basedtrajectory predictors. While these predictors are at the core of manyapplications such as autonomous navigation or pedestrian flow simulations,their adversarial robustness has not been carefully studied. In this paper, weintroduce a socially-attended attack to assess the social understanding ofprediction models in terms of collision avoidance. An attack is a small yetcarefully-crafted perturbations to fail predictors. Technically, we definecollision as a failure mode of the output, and propose hard- and soft-attentionmechanisms to guide our attack. Thanks to our attack, we shed light on thelimitations of the current models in terms of their social understanding. Wedemonstrate the strengths of our method on the recent trajectory predictionmodels. Finally, we show that our attack can be employed to increase the socialunderstanding of state-of-the-art models. The code is available online:https://s-attack.github.io/

Are socially-aware trajectory prediction models really socially-aware?