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/

Zendy Plus

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

Zendy Tools

Zendy Open

We propose a concise representation of videos that encode perceptuallymeaningful features into graphs. With this representation, we aim to leveragethe large amount of redundancies in videos and save computations. First, weconstruct superpixel-based graph representations of videos by consideringsuperpixels as graph nodes and create spatial and temporal connections betweenadjacent superpixels. Then, we leverage Graph Convolutional Networks to processthis representation and predict the desired output. As a result, we are able totrain models with much fewer parameters, which translates into short trainingperiods and a reduction in computation resource requirements. A comprehensiveexperimental study on the publicly available datasets Kinetics-400 and Charadesshows that the proposed method is highly cost-effective and uses limitedcommodity hardware during training and inference. It reduces the computationalrequirements 10-fold while achieving results that are comparable tostate-of-the-art methods. We believe that the proposed approach is a promisingdirection that could open the door to solving video understanding moreefficiently and enable more resource limited users to thrive in this researchfield.

GraphVid: It Only Takes a Few Nodes to Understand a Video