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

We consider the problem of video snapshot compressive imaging (SCI), wheresequential high-speed frames are modulated by different masks and captured by asingle measurement. The underlying principle of reconstructing multi-frameimages from only one single measurement is to solve an ill-posed problem. Bycombining optimization algorithms and neural networks, deep unfolding networks(DUNs) score tremendous achievements in solving inverse problems. In thispaper, our proposed model is under the DUN framework and we propose a 3DConvolution-Transformer Mixture (CTM) module with a 3D efficient and scalableattention model plugged in, which helps fully learn the correlation betweentemporal and spatial dimensions by virtue of Transformer. To our bestknowledge, this is the first time that Transformer is employed to video SCIreconstruction. Besides, to further investigate the high-frequency informationduring the reconstruction process which are neglected in previous studies, weintroduce variance estimation characterizing the uncertainty on apixel-by-pixel basis. Extensive experimental results demonstrate that ourproposed method achieves state-of-the-art (SOTA) (with a 1.2dB gain in PSNRover previous SOTA algorithm) results. We will release the code.

Unfolding Framework with Prior of Convolution-Transformer Mixture and Uncertainty Estimation for Video Snapshot Compressive Imaging