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

Keyword spotting (KWS) on mobile devices generally requires a small memoryfootprint. However, most current models still maintain a large number ofparameters in order to ensure good performance. In this paper, we propose atemporally pooled attention module which can capture global features betterthan the AveragePool. Besides, we design a separable temporal convolutionnetwork which leverages depthwise separable and temporal convolution to reducethe number of parameter and calculations. Finally, taking advantage ofseparable temporal convolution and temporally pooled attention, a efficientneural network (ST-AttNet) is designed for KWS system. We evaluate the modelson the publicly available Google speech commands data sets V1. The number ofparameters of proposed model (48K) is 1/6 of state-of-the-art TC-ResNet14-1.5model (305K). The proposed model achieves a 96.6% accuracy, which is comparableto the TC-ResNet14-1.5 model (96.6%).

Separable Temporal Convolution plus Temporally Pooled Attention for Lightweight High-performance Keyword Spotting