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

Mobile-centric AI applications have high requirements for resource-efficiencyof model inference. Input filtering is a promising approach to eliminate theredundancy so as to reduce the cost of inference. Previous efforts havetailored effective solutions for many applications, but left two essentialquestions unanswered: (1) theoretical filterability of an inference workload toguide the application of input filtering techniques, thereby avoiding thetrial-and-error cost for resource-constrained mobile applications; (2) robustdiscriminability of feature embedding to allow input filtering to be widelyeffective for diverse inference tasks and input content. To answer them, wefirst formalize the input filtering problem and theoretically compare thehypothesis complexity of inference models and input filters to understand theoptimization potential. Then we propose the first end-to-end learnable inputfiltering framework that covers most state-of-the-art methods and surpassesthem in feature embedding with robust discriminability. We design and implementInFi that supports six input modalities and multiple mobile-centricdeployments. Comprehensive evaluations confirm our theoretical results and showthat InFi outperforms strong baselines in applicability, accuracy, andefficiency. InFi achieve 8.5x throughput and save 95% bandwidth, while keepingover 90% accuracy, for a video analytics application on mobile platforms.

InFi: End-to-End Learning to Filter Input for Resource-Efficiency in Mobile-Centric Inference