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

Convolutional Neural Network(CNN) has been widely used for image recognitionwith great success. However, there are a number of limitations of the currentCNN based image recognition paradigm. First, the receptive field of CNN isgenerally fixed, which limits its recognition capacity when the input image isvery large. Second, it lacks the computational scalability for dealing withimages with different sizes. Third, it is quite different from human visualsystem for image recognition, which involves both feadforward and recurrentproprocessing. This paper proposes a different paradigm of image recognition,which can take advantages of variable scales of the input images, has morecomputational scalabilities, and is more similar to image recognition by humanvisual system. It is based on recurrent neural network (RNN) defined on imagescale with an embeded base CNN, which is named Scale Recurrent NeuralNetwork(SRNN). This RNN based approach makes it easier to deal with images withvariable sizes, and allows us to borrow existing RNN techniques, such as LSTMand GRU, to further enhance the recognition accuracy. Our experiments show thatthe recognition accuracy of a base CNN can be significantly boosted using theproposed SRNN models. It also significantly outperforms the scale ensemblemethod, which integrate the results of performing CNN to the input image atdifferent scales, although the computational overhead of using SRNN isnegligible.

Image Recognition Using Scale Recurrent Neural Networks