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

Model compression techniques on Deep Neural Network (DNN) have been widelyacknowledged as an effective way to achieve acceleration on a variety ofplatforms, and DNN weight pruning is a straightforward and effective method.There are currently two mainstreams of pruning methods representing twoextremes of pruning regularity: non-structured, fine-grained pruning canachieve high sparsity and accuracy, but is not hardware friendly; structured,coarse-grained pruning exploits hardware-efficient structures in pruning, butsuffers from accuracy drop when the pruning rate is high. In this paper, weintroduce PCONV, comprising a new sparsity dimension, -- fine-grained pruningpatterns inside the coarse-grained structures. PCONV comprises two types ofsparsities, Sparse Convolution Patterns (SCP) which is generated fromintra-convolution kernel pruning and connectivity sparsity generated frominter-convolution kernel pruning. Essentially, SCP enhances accuracy due to itsspecial vision properties, and connectivity sparsity increases pruning ratewhile maintaining balanced workload on filter computation. To deploy PCONV, wedevelop a novel compiler-assisted DNN inference framework and execute PCONVmodels in real-time without accuracy compromise, which cannot be achieved inprior work. Our experimental results show that, PCONV outperforms threestate-of-art end-to-end DNN frameworks, TensorFlow-Lite, TVM, and AlibabaMobile Neural Network with speedup up to 39.2x, 11.4x, and 6.3x, respectively,with no accuracy loss. Mobile devices can achieve real-time inference onlarge-scale DNNs.

PCONV: The Missing but Desirable Sparsity in DNN Weight Pruning for Real-time Execution on Mobile Devices