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

The inception network has been shown to provide good performance on imageclassification problems, but there are not much evidences that it is alsoeffective for the image restoration or pixel-wise labeling problems. For imagerestoration problems, the pooling is generally not used because the decimatedfeatures are not helpful for the reconstruction of an image as the output.Moreover, most deep learning architectures for the restoration problems do notuse dense prediction that need lots of training parameters. From theseobservations, for enjoying the performance of inception-like structure on theimage based problems we propose a new convolutional network-in-networkstructure. The proposed network can be considered a modification of inceptionstructure where pool projection and pooling layer are removed for maintainingthe entire feature map size, and a larger kernel filter is added instead.Proposed network greatly reduces the number of parameters on account of removeddense prediction and pooling, which is an advantage, but may also reduce thereceptive field in each layer. Hence, we add a larger kernel than the originalinception structure for not increasing the depth of layers. The proposedstructure is applied to typical image-to-image learning problems, i.e., theproblems where the size of input and output are same such as skin detection,semantic segmentation, and compression artifacts reduction. Extensiveexperiments show that the proposed network brings comparable or better resultsthan the state-of-the-art convolutional neural networks for these problems.

A New Convolutional Network-in-Network Structure and Its Applications in Skin Detection, Semantic Segmentation, and Artifact Reduction