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

Fast model updates for unseen tasks on intelligent edge devices are crucialbut also challenging due to the limited computational power. In this paper,wepropose MetaLDC, which meta-trains braininspired ultra-efficientlow-dimensional computing classifiers to enable fast adaptation on tiny deviceswith minimal computational costs. Concretely, during the meta-training stage,MetaLDC meta trains a representation offline by explicitly taking into accountthat the final (binary) class layer will be fine-tuned for fast adaptation forunseen tasks on tiny devices; during the meta-testing stage, MetaLDC usesclosed-form gradients of the loss function to enable fast adaptation of theclass layer. Unlike traditional neural networks, MetaLDC is designed based onthe emerging LDC framework to enable ultra-efficient on-device inference. Ourexperiments have demonstrated that compared to SOTA baselines, MetaLDC achieveshigher accuracy, robustness against random bit errors, as well ascost-efficient hardware computation.

MetaLDC: Meta Learning of Low-Dimensional Computing Classifiers for Fast On-Device Adaption