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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Among applications of deep learning (DL) involving low cost sensors, remoteimage classification involves a physical channel that separates edge sensorsand cloud classifiers. Traditional DL models must be divided between an encoderfor the sensor and the decoder + classifier at the edge server. An importantchallenge is to effectively train such distributed models when the connectingchannels have limited rate/capacity. Our goal is to optimize DL models suchthat the encoder latent requires low channel bandwidth while still deliversfeature information for high classification accuracy. This work proposes athree-step joint learning strategy to guide encoders to extract features thatare compact, discriminative, and amenable to commonaugmentations/transformations. We optimize latent dimension through an initialscreening phase before end-to-end (E2E) training. To obtain an adjustable bitrate via a single pre-deployed encoder, we apply entropy-based quantizationand/or manual truncation on the latent representations. Tests show that ourproposed method achieves accuracy improvement of up to 1.5% on CIFAR-10 and 3%on CIFAR-100 over conventional E2E cross-entropy training.

A Principled Hierarchical Deep Learning Approach to Joint Image Compression and Classification