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

Applications of surveillance video have developed rapidly in recent years toprotect public safety and daily life, which often detect and recognize objectsin video sequences. Traditional coding frameworks remove temporal redundancy insurveillance video by block-wise motion compensation, lacking the extractionand utilization of inherent structure information. In this paper, we figure outthis issue by disentangling surveillance video into the structure of a globalspatio-temporal feature (memory) for Group of Picture (GoP) and skeleton foreach frame (clue). The memory is obtained by sequentially feeding frame insideGoP into a recurrent neural network, describing appearance for objects thatappeared inside GoP. While the skeleton is calculated by a pose estimator, itis regarded as a clue to recall memory. Furthermore, an attention mechanism isintroduced to obtain the relation between appearance and skeletons. Finally, weemploy generative adversarial network to reconstruct each frame. Experimentalresults indicate that our method effectively generates realistic reconstructionbased on appearance and skeleton, which show much higher compressionperformance on surveillance video compared with the latest video compressionstandard H.265.

Generative Memorize-Then-Recall framework for low bit-rate Surveillance Video Compression