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

Precisely naming the action depicted in a video can be a challenging andoftentimes ambiguous task. In contrast to object instances represented as nouns(e.g. dog, cat, chair, etc.), in the case of actions, human annotatorstypically lack a consensus as to what constitutes a specific action (e.g.jogging versus running). In practice, a given video can contain multiple validpositive annotations for the same action. As a result, video datasets oftencontain significant levels of label noise and overlap between the atomic actionclasses. In this work, we address the challenge of training multi-label actionrecognition models from only single positive training labels. We propose twoapproaches that are based on generating pseudo training examples sampled fromsimilar instances within the train set. Unlike other approaches that usemodel-derived pseudo-labels, our pseudo-labels come from human annotations andare selected based on feature similarity. To validate our approaches, we createa new evaluation benchmark by manually annotating a subset ofEPIC-Kitchens-100's validation set with multiple verb labels. We presentresults on this new test set along with additional results on a new version ofHMDB-51, called Confusing-HMDB-102, where we outperform existing methods inboth cases. Data and code are available athttps://github.com/kiyoon/verb_ambiguity

An Action Is Worth Multiple Words: Handling Ambiguity in Action Recognition