Matches in SemOpenAlex for { <https://semopenalex.org/work/W2914277910> ?p ?o ?g. }
Showing items 1 to 74 of
74
with 100 items per page.
- W2914277910 abstract "In low-resource scenarios, for example, small datasets or a lack in computational resources available, state-of-the-art deep learning methods for speech recognition have been known to fail. It is possible to achieve more robust models if care is taken to ensure the learning guarantees provided by the statistical learning theory. This work presents a shallow and hybrid approach using a convolutional neural network feature extractor fed into a hierarchical tree of support vector machines for classification. Here, we show that gross errors present even in state-of-the-art systems can be avoided and that an accurate acoustic model can be built in a hierarchical fashion. Furthermore, we present proof that our algorithm does adhere to the learning guarantees provided by the statistical learning theory. The acoustic model produced in this work outperforms traditional hidden Markov models, and the hierarchical support vector machine tree outperforms a multi-class multilayer perceptron classifier using the same features. More importantly, we isolate the performance of the acoustic model and provide results on both the frame and phoneme level, considering the true robustness of the model. We show that even with a small amount of data, accurate and robust recognition rates can be obtained." @default.
- W2914277910 created "2019-02-21" @default.
- W2914277910 creator A5032486030 @default.
- W2914277910 creator A5057570319 @default.
- W2914277910 creator A5059202822 @default.
- W2914277910 creator A5073145218 @default.
- W2914277910 date "2019-01-04" @default.
- W2914277910 modified "2023-09-23" @default.
- W2914277910 title "Theoretical learning guarantees applied to acoustic modeling" @default.
- W2914277910 cites W1218987319 @default.
- W2914277910 cites W1494198834 @default.
- W2914277910 cites W1608367484 @default.
- W2914277910 cites W1993882792 @default.
- W2914277910 cites W1995562189 @default.
- W2914277910 cites W2005708641 @default.
- W2914277910 cites W2024119716 @default.
- W2914277910 cites W2029996593 @default.
- W2914277910 cites W2077804127 @default.
- W2914277910 cites W2112739286 @default.
- W2914277910 cites W2112796928 @default.
- W2914277910 cites W2116360511 @default.
- W2914277910 cites W2117671523 @default.
- W2914277910 cites W2125496931 @default.
- W2914277910 cites W2131700150 @default.
- W2914277910 cites W2148143831 @default.
- W2914277910 cites W2155273149 @default.
- W2914277910 cites W2160815625 @default.
- W2914277910 cites W2296265275 @default.
- W2914277910 cites W2299809954 @default.
- W2914277910 cites W2552635739 @default.
- W2914277910 cites W2765992979 @default.
- W2914277910 cites W2766439643 @default.
- W2914277910 cites W2919115771 @default.
- W2914277910 cites W2962738191 @default.
- W2914277910 cites W3022436500 @default.
- W2914277910 cites W4299847517 @default.
- W2914277910 doi "https://doi.org/10.1186/s13173-018-0081-3" @default.
- W2914277910 hasPublicationYear "2019" @default.
- W2914277910 type Work @default.
- W2914277910 sameAs 2914277910 @default.
- W2914277910 citedByCount "6" @default.
- W2914277910 countsByYear W29142779102019 @default.
- W2914277910 countsByYear W29142779102020 @default.
- W2914277910 countsByYear W29142779102022 @default.
- W2914277910 crossrefType "journal-article" @default.
- W2914277910 hasAuthorship W2914277910A5032486030 @default.
- W2914277910 hasAuthorship W2914277910A5057570319 @default.
- W2914277910 hasAuthorship W2914277910A5059202822 @default.
- W2914277910 hasAuthorship W2914277910A5073145218 @default.
- W2914277910 hasBestOaLocation W29142779101 @default.
- W2914277910 hasConcept C154945302 @default.
- W2914277910 hasConcept C41008148 @default.
- W2914277910 hasConceptScore W2914277910C154945302 @default.
- W2914277910 hasConceptScore W2914277910C41008148 @default.
- W2914277910 hasIssue "1" @default.
- W2914277910 hasLocation W29142779101 @default.
- W2914277910 hasLocation W29142779102 @default.
- W2914277910 hasOpenAccess W2914277910 @default.
- W2914277910 hasPrimaryLocation W29142779101 @default.
- W2914277910 hasRelatedWork W2049775471 @default.
- W2914277910 hasRelatedWork W2093578348 @default.
- W2914277910 hasRelatedWork W2350741829 @default.
- W2914277910 hasRelatedWork W2358668433 @default.
- W2914277910 hasRelatedWork W2376932109 @default.
- W2914277910 hasRelatedWork W2382290278 @default.
- W2914277910 hasRelatedWork W2390279801 @default.
- W2914277910 hasRelatedWork W2748952813 @default.
- W2914277910 hasRelatedWork W2899084033 @default.
- W2914277910 hasRelatedWork W3004735627 @default.
- W2914277910 hasVolume "25" @default.
- W2914277910 isParatext "false" @default.
- W2914277910 isRetracted "false" @default.
- W2914277910 magId "2914277910" @default.
- W2914277910 workType "article" @default.