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

Recent innovations in training deep convolutional neural network (ConvNet)models have motivated the design of new methods to automatically learn localimage descriptors. The latest deep ConvNets proposed for this task consist of asiamese network that is trained by penalising misclassification of pairs oflocal image patches. Current results from machine learning show that replacingthis siamese by a triplet network can improve the classification accuracy inseveral problems, but this has yet to be demonstrated for local imagedescriptor learning. Moreover, current siamese and triplet networks have beentrained with stochastic gradient descent that computes the gradient fromindividual pairs or triplets of local image patches, which can make them proneto overfitting. In this paper, we first propose the use of triplet networks forthe problem of local image descriptor learning. Furthermore, we also proposethe use of a global loss that minimises the overall classification error in thetraining set, which can improve the generalisation capability of the model.Using the UBC benchmark dataset for comparing local image descriptors, we showthat the triplet network produces a more accurate embedding than the siamesenetwork in terms of the UBC dataset errors. Moreover, we also demonstrate thata combination of the triplet and global losses produces the best embedding inthe field, using this triplet network. Finally, we also show that the use ofthe central-surround siamese network trained with the global loss produces thebest result of the field on the UBC dataset. Pre-trained models are availableonline at https://github.com/vijaykbg/deep-patchmatch

Learning Local Image Descriptors with Deep Siamese and Triplet Convolutional Networks by Minimising Global Loss Functions