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

Objects are crucial for understanding human-object interactions. Byidentifying the relevant objects, one can also predict potential futureinteractions or actions that may occur with these objects. In this paper, westudy the problem of Short-Term Object interaction anticipation (STA) andpropose NAOGAT (Next-Active-Object Guided Anticipation Transformer), amulti-modal end-to-end transformer network, that attends to objects in observedframes in order to anticipate the next-active-object (NAO) and, eventually, toguide the model to predict context-aware future actions. The task ischallenging since it requires anticipating future action along with the objectwith which the action occurs and the time after which the interaction willbegin, a.k.a. the time to contact (TTC). Compared to existing video modelingarchitectures for action anticipation, NAOGAT captures the relationship betweenobjects and the global scene context in order to predict detections for thenext active object and anticipate relevant future actions given thesedetections, leveraging the objects' dynamics to improve accuracy. One of thekey strengths of our approach, in fact, is its ability to exploit the motiondynamics of objects within a given clip, which is often ignored by othermodels, and separately decoding the object-centric and motion-centricinformation. Through our experiments, we show that our model outperformsexisting methods on two separate datasets, Ego4D and EpicKitchens-100 ("UnseenSet"), as measured by several additional metrics, such as time to contact, andnext-active-object localization. The code will be available upon acceptance.

Leveraging Next-Active Objects for Context-Aware Anticipation in Egocentric Videos