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

In image classification, visual separability between different objectcategories is highly uneven, and some categories are more difficult todistinguish than others. Such difficult categories demand more dedicatedclassifiers. However, existing deep convolutional neural networks (CNN) aretrained as flat N-way classifiers, and few efforts have been made to leveragethe hierarchical structure of categories. In this paper, we introducehierarchical deep CNNs (HD-CNNs) by embedding deep CNNs into a categoryhierarchy. An HD-CNN separates easy classes using a coarse category classifierwhile distinguishing difficult classes using fine category classifiers. DuringHD-CNN training, component-wise pretraining is followed by global finetuningwith a multinomial logistic loss regularized by a coarse category consistencyterm. In addition, conditional executions of fine category classifiers andlayer parameter compression make HD-CNNs scalable for large-scale visualrecognition. We achieve state-of-the-art results on both CIFAR100 andlarge-scale ImageNet 1000-class benchmark datasets. In our experiments, webuild up three different HD-CNNs and they lower the top-1 error of the standardCNNs by 2.65%, 3.1% and 1.1%, respectively.

HD-CNN: Hierarchical Deep Convolutional Neural Network for Large Scale Visual Recognition