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

Recently, much progress has been made for self-supervised action recognition.Most existing approaches emphasize the contrastive relations among videos,including appearance and motion consistency. However, two main issues remainfor existing pre-training methods: 1) the learned representation is neutral andnot informative for a specific task; 2) multi-task learning-based pre-trainingsometimes leads to sub-optimal solutions due to inconsistent domains ofdifferent tasks. To address the above issues, we propose a novel actionrecognition pre-training framework, which exploits human-centered priorknowledge that generates more informative representation, and avoids theconflict between multiple tasks by using task-dependent representations.Specifically, we distill knowledge from a human parsing model to enrich thesemantic capability of representation. In addition, we combine knowledgedistillation with contrastive learning to constitute a task-dependentmulti-task framework. We achieve state-of-the-art performance on two popularbenchmarks for action recognition task, i.e., UCF101 and HMDB51, verifying theeffectiveness of our method.

Human-Centered Prior-Guided and Task-Dependent Multi-Task Representation Learning for Action Recognition Pre-Training