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

This paper addresses the real-time encoding-decoding problem forhigh-frame-rate video compressive sensing (CS). Unlike prior works that performreconstruction using iterative optimization-based approaches, we propose anon-iterative model, named "CSVideoNet". CSVideoNet directly learns the inversemapping of CS and reconstructs the original input in a single forwardpropagation. To overcome the limitations of existing CS cameras, we propose amulti-rate CNN and a synthesizing RNN to improve the trade-off betweencompression ratio (CR) and spatial-temporal resolution of the reconstructedvideos. The experiment results demonstrate that CSVideoNet significantlyoutperforms the state-of-the-art approaches. With no pre/post-processing, weachieve 25dB PSNR recovery quality at 100x CR, with a frame rate of 125 fps ona Titan X GPU. Due to the feedforward and high-data-concurrency natures ofCSVideoNet, it can take advantage of GPU acceleration to achieve three ordersof magnitude speed-up over conventional iterative-based approaches. We sharethe source code at https://github.com/PSCLab-ASU/CSVideoNet.

CSVideoNet: A Real-time End-to-end Learning Framework for High-frame-rate Video Compressive Sensing