Matches in SemOpenAlex for { <https://semopenalex.org/work/W3115583060> ?p ?o ?g. }
Showing items 1 to 98 of
98
with 100 items per page.
- W3115583060 abstract "This paper introduces neural architecture search (NAS) for the automatic discovery of small models for keyword spotting (KWS) in limited resource environments. We employ a differentiable NAS approach to optimize the structure of convolutional neural networks (CNNs) to maximize the classification accuracy while minimizing the number of operations per inference. Using NAS only, we were able to obtain a highly efficient model with 95.4% accuracy on the Google speech commands dataset with 494.8 kB of memory usage and 19.6 million operations. Additionally, weight quantization is used to reduce the memory consumption even further. We show that weight quantization to low bit-widths (e.g. 1 bit) can be used without substantial loss in accuracy. By increasing the number of input features from 10 MFCC to 20 MFCC we were able to increase the accuracy to 96.3% at 340.1 kB of memory usage and 27.1 million operations." @default.
- W3115583060 created "2021-01-05" @default.
- W3115583060 creator A5015798259 @default.
- W3115583060 creator A5017995708 @default.
- W3115583060 creator A5019416885 @default.
- W3115583060 date "2020-12-18" @default.
- W3115583060 modified "2023-09-27" @default.
- W3115583060 title "Resource-efficient DNNs for Keyword Spotting using Neural Architecture Search and Quantization" @default.
- W3115583060 cites W1665214252 @default.
- W3115583060 cites W1836465849 @default.
- W3115583060 cites W2034940213 @default.
- W3115583060 cites W2119717200 @default.
- W3115583060 cites W2242818861 @default.
- W3115583060 cites W2267635276 @default.
- W3115583060 cites W2276486856 @default.
- W3115583060 cites W2291160084 @default.
- W3115583060 cites W2294282016 @default.
- W3115583060 cites W2469490737 @default.
- W3115583060 cites W2595614461 @default.
- W3115583060 cites W2613332842 @default.
- W3115583060 cites W2769912137 @default.
- W3115583060 cites W2797583228 @default.
- W3115583060 cites W2810075754 @default.
- W3115583060 cites W2888429796 @default.
- W3115583060 cites W2953219395 @default.
- W3115583060 cites W2955425717 @default.
- W3115583060 cites W2962746461 @default.
- W3115583060 cites W2963163009 @default.
- W3115583060 cites W2963473542 @default.
- W3115583060 cites W2963778169 @default.
- W3115583060 cites W2963821229 @default.
- W3115583060 cites W2963918968 @default.
- W3115583060 cites W2964081807 @default.
- W3115583060 cites W2964259004 @default.
- W3115583060 cites W2965658867 @default.
- W3115583060 cites W2999963857 @default.
- W3115583060 cites W3098773154 @default.
- W3115583060 cites W2602634800 @default.
- W3115583060 cites W3022571137 @default.
- W3115583060 hasPublicationYear "2020" @default.
- W3115583060 type Work @default.
- W3115583060 sameAs 3115583060 @default.
- W3115583060 citedByCount "0" @default.
- W3115583060 crossrefType "posted-content" @default.
- W3115583060 hasAuthorship W3115583060A5015798259 @default.
- W3115583060 hasAuthorship W3115583060A5017995708 @default.
- W3115583060 hasAuthorship W3115583060A5019416885 @default.
- W3115583060 hasConcept C11413529 @default.
- W3115583060 hasConcept C119857082 @default.
- W3115583060 hasConcept C153180895 @default.
- W3115583060 hasConcept C154945302 @default.
- W3115583060 hasConcept C2776214188 @default.
- W3115583060 hasConcept C2779506182 @default.
- W3115583060 hasConcept C2781213101 @default.
- W3115583060 hasConcept C28490314 @default.
- W3115583060 hasConcept C28855332 @default.
- W3115583060 hasConcept C41008148 @default.
- W3115583060 hasConcept C50644808 @default.
- W3115583060 hasConcept C81363708 @default.
- W3115583060 hasConceptScore W3115583060C11413529 @default.
- W3115583060 hasConceptScore W3115583060C119857082 @default.
- W3115583060 hasConceptScore W3115583060C153180895 @default.
- W3115583060 hasConceptScore W3115583060C154945302 @default.
- W3115583060 hasConceptScore W3115583060C2776214188 @default.
- W3115583060 hasConceptScore W3115583060C2779506182 @default.
- W3115583060 hasConceptScore W3115583060C2781213101 @default.
- W3115583060 hasConceptScore W3115583060C28490314 @default.
- W3115583060 hasConceptScore W3115583060C28855332 @default.
- W3115583060 hasConceptScore W3115583060C41008148 @default.
- W3115583060 hasConceptScore W3115583060C50644808 @default.
- W3115583060 hasConceptScore W3115583060C81363708 @default.
- W3115583060 hasLocation W31155830601 @default.
- W3115583060 hasOpenAccess W3115583060 @default.
- W3115583060 hasPrimaryLocation W31155830601 @default.
- W3115583060 hasRelatedWork W2534720278 @default.
- W3115583060 hasRelatedWork W2781820017 @default.
- W3115583060 hasRelatedWork W2793950911 @default.
- W3115583060 hasRelatedWork W2904036992 @default.
- W3115583060 hasRelatedWork W2905687271 @default.
- W3115583060 hasRelatedWork W2996159591 @default.
- W3115583060 hasRelatedWork W3003530529 @default.
- W3115583060 hasRelatedWork W3006769269 @default.
- W3115583060 hasRelatedWork W3082696729 @default.
- W3115583060 hasRelatedWork W3091564954 @default.
- W3115583060 hasRelatedWork W3097018422 @default.
- W3115583060 hasRelatedWork W3098124779 @default.
- W3115583060 hasRelatedWork W3105075025 @default.
- W3115583060 hasRelatedWork W3118322613 @default.
- W3115583060 hasRelatedWork W3134145319 @default.
- W3115583060 hasRelatedWork W3159027058 @default.
- W3115583060 hasRelatedWork W3169638910 @default.
- W3115583060 hasRelatedWork W3178567251 @default.
- W3115583060 hasRelatedWork W3190530808 @default.
- W3115583060 hasRelatedWork W3198685759 @default.
- W3115583060 isParatext "false" @default.
- W3115583060 isRetracted "false" @default.
- W3115583060 magId "3115583060" @default.
- W3115583060 workType "article" @default.