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 address the problem of removing undesirable reflections from a singleimage captured through a glass surface, which is an ill-posed, challenging butpractically important problem for photo enhancement. Inspired by iterativestructure reduction for hidden community detection in social networks, wepropose an Iterative Boost Convolutional LSTM Network (IBCLN) that enablescascaded prediction for reflection removal. IBCLN is a cascaded network thatiteratively refines the estimates of transmission and reflection layers in amanner that they can boost the prediction quality to each other, andinformation across steps of the cascade is transferred using an LSTM. Theintuition is that the transmission is the strong, dominant structure while thereflection is the weak, hidden structure. They are complementary to each otherin a single image and thus a better estimate and reduction on one side from theoriginal image leads to a more accurate estimate on the other side. Tofacilitate training over multiple cascade steps, we employ LSTM to address thevanishing gradient problem, and propose residual reconstruction loss as furthertraining guidance. Besides, we create a dataset of real-world images withreflection and ground-truth transmission layers to mitigate the problem ofinsufficient data. Comprehensive experiments demonstrate that the proposedmethod can effectively remove reflections in real and synthetic images comparedwith state-of-the-art reflection removal methods.

Single Image Reflection Removal through Cascaded Refinement