Matches in SemOpenAlex for { <https://semopenalex.org/work/W4378373450> ?p ?o ?g. }
- W4378373450 endingPage "53236" @default.
- W4378373450 startingPage "53224" @default.
- W4378373450 abstract "Keyword Spotting (KWS) has been the subject of research in recent years given the increase of embedded systems for command recognition such as Alexa, Google Home, and Siri. Performance, model size, processing time, and robustness to noise are fundamental in these systems. Furthermore, applications in embedded systems demand computationally efficient models that can be implemented in current technology. In this work, an approach for keyword recognition is evaluated using three deep learning models namely LeNet-5, SqueezeNet, and EfficientNet-B0. We evaluate transfer learning, pruning and quantization strategies in training and test using noisy and clean speech signals. In addition, compression techniques such as pruning and quantization were assessed in terms of the size reduction of the model footprint and the accuracy obtained in each case. Using the Google’s Speech Commands dataset and additive babble noise signal, our keyword recognition approach achieves an accuracy of 94.6% using an unstructured pruning of 80% of the parameters of the original SqueezeNet network with a reduction of 70% in the model size." @default.
- W4378373450 created "2023-05-27" @default.
- W4378373450 creator A5021252790 @default.
- W4378373450 creator A5024992483 @default.
- W4378373450 creator A5067575855 @default.
- W4378373450 date "2023-01-01" @default.
- W4378373450 modified "2023-10-10" @default.
- W4378373450 title "Evaluating Robustness to Noise and Compression of Deep Neural Networks for Keyword Spotting" @default.
- W4378373450 cites W1974387177 @default.
- W4378373450 cites W2034940213 @default.
- W4378373450 cites W2060108852 @default.
- W4378373450 cites W2092443213 @default.
- W4378373450 cites W2097117768 @default.
- W4378373450 cites W2103869314 @default.
- W4378373450 cites W2107745473 @default.
- W4378373450 cites W2112796928 @default.
- W4378373450 cites W2117633163 @default.
- W4378373450 cites W2125838338 @default.
- W4378373450 cites W2157562541 @default.
- W4378373450 cites W2191779130 @default.
- W4378373450 cites W2294370754 @default.
- W4378373450 cites W2407023693 @default.
- W4378373450 cites W2507580616 @default.
- W4378373450 cites W2886352393 @default.
- W4378373450 cites W2888728157 @default.
- W4378373450 cites W2913485229 @default.
- W4378373450 cites W2963446712 @default.
- W4378373450 cites W3015287265 @default.
- W4378373450 cites W3038838661 @default.
- W4378373450 cites W3094272606 @default.
- W4378373450 cites W3097018422 @default.
- W4378373450 cites W3103436470 @default.
- W4378373450 cites W3154934206 @default.
- W4378373450 cites W3163571828 @default.
- W4378373450 cites W3176579201 @default.
- W4378373450 cites W3184606595 @default.
- W4378373450 cites W3196496149 @default.
- W4378373450 cites W3197949759 @default.
- W4378373450 cites W3198035615 @default.
- W4378373450 cites W3199053851 @default.
- W4378373450 cites W4200323562 @default.
- W4378373450 cites W4206567542 @default.
- W4378373450 cites W4210849193 @default.
- W4378373450 cites W4226253207 @default.
- W4378373450 cites W4236937157 @default.
- W4378373450 cites W4294310677 @default.
- W4378373450 cites W4296708560 @default.
- W4378373450 cites W4313178214 @default.
- W4378373450 doi "https://doi.org/10.1109/access.2023.3280477" @default.
- W4378373450 hasPublicationYear "2023" @default.
- W4378373450 type Work @default.
- W4378373450 citedByCount "0" @default.
- W4378373450 crossrefType "journal-article" @default.
- W4378373450 hasAuthorship W4378373450A5021252790 @default.
- W4378373450 hasAuthorship W4378373450A5024992483 @default.
- W4378373450 hasAuthorship W4378373450A5067575855 @default.
- W4378373450 hasBestOaLocation W43783734501 @default.
- W4378373450 hasConcept C104317684 @default.
- W4378373450 hasConcept C108583219 @default.
- W4378373450 hasConcept C119857082 @default.
- W4378373450 hasConcept C153180895 @default.
- W4378373450 hasConcept C154945302 @default.
- W4378373450 hasConcept C185592680 @default.
- W4378373450 hasConcept C2781213101 @default.
- W4378373450 hasConcept C28490314 @default.
- W4378373450 hasConcept C28855332 @default.
- W4378373450 hasConcept C31972630 @default.
- W4378373450 hasConcept C41008148 @default.
- W4378373450 hasConcept C50644808 @default.
- W4378373450 hasConcept C55493867 @default.
- W4378373450 hasConcept C63479239 @default.
- W4378373450 hasConceptScore W4378373450C104317684 @default.
- W4378373450 hasConceptScore W4378373450C108583219 @default.
- W4378373450 hasConceptScore W4378373450C119857082 @default.
- W4378373450 hasConceptScore W4378373450C153180895 @default.
- W4378373450 hasConceptScore W4378373450C154945302 @default.
- W4378373450 hasConceptScore W4378373450C185592680 @default.
- W4378373450 hasConceptScore W4378373450C2781213101 @default.
- W4378373450 hasConceptScore W4378373450C28490314 @default.
- W4378373450 hasConceptScore W4378373450C28855332 @default.
- W4378373450 hasConceptScore W4378373450C31972630 @default.
- W4378373450 hasConceptScore W4378373450C41008148 @default.
- W4378373450 hasConceptScore W4378373450C50644808 @default.
- W4378373450 hasConceptScore W4378373450C55493867 @default.
- W4378373450 hasConceptScore W4378373450C63479239 @default.
- W4378373450 hasFunder F4320320997 @default.
- W4378373450 hasFunder F4320321091 @default.
- W4378373450 hasLocation W43783734501 @default.
- W4378373450 hasOpenAccess W4378373450 @default.
- W4378373450 hasPrimaryLocation W43783734501 @default.
- W4378373450 hasRelatedWork W2795261237 @default.
- W4378373450 hasRelatedWork W3014300295 @default.
- W4378373450 hasRelatedWork W3164822677 @default.
- W4378373450 hasRelatedWork W4223943233 @default.
- W4378373450 hasRelatedWork W4225161397 @default.
- W4378373450 hasRelatedWork W4312200629 @default.
- W4378373450 hasRelatedWork W4360585206 @default.
- W4378373450 hasRelatedWork W4364306694 @default.