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

Convolution Neural Networks (CNN) have performed well in many applicationssuch as object detection, pattern recognition, video surveillance and so on.CNN carryout feature extraction on labelled data to perform classification.Multi-label classification assigns more than one label to a particular datasample in a data set. In multi-label classification, properties of a data pointthat are considered to be mutually exclusive are classified. However, existingmulti-label classification requires some form of data pre-processing thatinvolves image training data cropping or image tiling. The computation andmemory requirement of these multi-label CNN models makes their deployment onedge devices challenging. In this paper, we propose a methodology that solvesthis problem by extending the capability of existing multi-label classificationand provide models with lower latency that requires smaller memory size whendeployed on edge devices. We make use of a single CNN model designed withmultiple loss layers and multiple accuracy layers. This methodology is testedon state-of-the-art deep learning algorithms such as AlexNet, GoogleNet andSqueezeNet using the Stanford Cars Dataset and deployed on Raspberry Pi3. Fromthe results the proposed methodology achieves comparable accuracy with 1.8xless MACC operation, 0.97x reduction in latency and 0.5x, 0.84x and 0.97xreduction in size for the generated AlexNet, GoogleNet and SqueezeNet CNNmodels respectively when compared to conventional ways of achieving multi-labelclassification like hard-coding multi-label instances into single labels. Themethodology also yields CNN models that achieve 50\% less MACC operations, 50%reduction in latency and size of generated versions of AlexNet, GoogleNet andSqueezeNet respectively when compared to conventional ways using 2 differentsingle-labelled models to achieve multi-label classification.

A Scalable Multilabel Classification to Deploy Deep Learning Architectures For Edge Devices