Matches in SemOpenAlex for { <https://semopenalex.org/work/W3016188195> ?p ?o ?g. }
Showing items 1 to 86 of
86
with 100 items per page.
- W3016188195 abstract "End-to-end approaches for automatic speech recognition (ASR) benefit from directly modeling the probability of the word sequence given the input audio stream in a single neural network. However, compared to conventional ASR systems, these models typically require more data to achieve comparable results. Well-known model adaptation techniques, to account for domain and style adaptation, are not easily applicable to end-to-end systems. Conventional HMM-based systems, on the other hand, have been optimized for various production environments and use cases. In this work, we propose to combine the benefits of end-to-end approaches with a conventional system using an attention-based discriminative language model that learns to rescore the output of a first-pass ASR system. We show that learning to rescore a list of potential ASR outputs is much simpler than learning to generate the hypothesis. The proposed model results in up to 8% improvement in word error rate even when the amount of training data is a fraction of data used for training the first-pass system." @default.
- W3016188195 created "2020-04-17" @default.
- W3016188195 creator A5008201430 @default.
- W3016188195 creator A5014054572 @default.
- W3016188195 date "2020-05-01" @default.
- W3016188195 modified "2023-10-16" @default.
- W3016188195 title "Audio-Attention Discriminative Language Model for ASR Rescoring" @default.
- W3016188195 cites W1600744878 @default.
- W3016188195 cites W1934041838 @default.
- W3016188195 cites W2026149468 @default.
- W3016188195 cites W2064675550 @default.
- W3016188195 cites W2143612262 @default.
- W3016188195 cites W2160815625 @default.
- W3016188195 cites W2396464458 @default.
- W3016188195 cites W2622203030 @default.
- W3016188195 cites W2750499125 @default.
- W3016188195 cites W2886319145 @default.
- W3016188195 cites W2916997151 @default.
- W3016188195 cites W2937402758 @default.
- W3016188195 cites W2962824709 @default.
- W3016188195 cites W2963660642 @default.
- W3016188195 cites W2972320704 @default.
- W3016188195 cites W2972625221 @default.
- W3016188195 doi "https://doi.org/10.1109/icassp40776.2020.9054335" @default.
- W3016188195 hasPublicationYear "2020" @default.
- W3016188195 type Work @default.
- W3016188195 sameAs 3016188195 @default.
- W3016188195 citedByCount "16" @default.
- W3016188195 countsByYear W30161881952020 @default.
- W3016188195 countsByYear W30161881952021 @default.
- W3016188195 countsByYear W30161881952022 @default.
- W3016188195 countsByYear W30161881952023 @default.
- W3016188195 crossrefType "proceedings-article" @default.
- W3016188195 hasAuthorship W3016188195A5008201430 @default.
- W3016188195 hasAuthorship W3016188195A5014054572 @default.
- W3016188195 hasBestOaLocation W30161881952 @default.
- W3016188195 hasConcept C119857082 @default.
- W3016188195 hasConcept C120665830 @default.
- W3016188195 hasConcept C121332964 @default.
- W3016188195 hasConcept C137293760 @default.
- W3016188195 hasConcept C138885662 @default.
- W3016188195 hasConcept C139807058 @default.
- W3016188195 hasConcept C154945302 @default.
- W3016188195 hasConcept C23224414 @default.
- W3016188195 hasConcept C28490314 @default.
- W3016188195 hasConcept C40969351 @default.
- W3016188195 hasConcept C41008148 @default.
- W3016188195 hasConcept C41895202 @default.
- W3016188195 hasConcept C50644808 @default.
- W3016188195 hasConcept C74296488 @default.
- W3016188195 hasConcept C90805587 @default.
- W3016188195 hasConcept C97931131 @default.
- W3016188195 hasConceptScore W3016188195C119857082 @default.
- W3016188195 hasConceptScore W3016188195C120665830 @default.
- W3016188195 hasConceptScore W3016188195C121332964 @default.
- W3016188195 hasConceptScore W3016188195C137293760 @default.
- W3016188195 hasConceptScore W3016188195C138885662 @default.
- W3016188195 hasConceptScore W3016188195C139807058 @default.
- W3016188195 hasConceptScore W3016188195C154945302 @default.
- W3016188195 hasConceptScore W3016188195C23224414 @default.
- W3016188195 hasConceptScore W3016188195C28490314 @default.
- W3016188195 hasConceptScore W3016188195C40969351 @default.
- W3016188195 hasConceptScore W3016188195C41008148 @default.
- W3016188195 hasConceptScore W3016188195C41895202 @default.
- W3016188195 hasConceptScore W3016188195C50644808 @default.
- W3016188195 hasConceptScore W3016188195C74296488 @default.
- W3016188195 hasConceptScore W3016188195C90805587 @default.
- W3016188195 hasConceptScore W3016188195C97931131 @default.
- W3016188195 hasLocation W30161881951 @default.
- W3016188195 hasLocation W30161881952 @default.
- W3016188195 hasOpenAccess W3016188195 @default.
- W3016188195 hasPrimaryLocation W30161881951 @default.
- W3016188195 hasRelatedWork W1491120356 @default.
- W3016188195 hasRelatedWork W1574295218 @default.
- W3016188195 hasRelatedWork W2048917925 @default.
- W3016188195 hasRelatedWork W2106051978 @default.
- W3016188195 hasRelatedWork W2124094254 @default.
- W3016188195 hasRelatedWork W2167513360 @default.
- W3016188195 hasRelatedWork W2259397180 @default.
- W3016188195 hasRelatedWork W2608712415 @default.
- W3016188195 hasRelatedWork W2996122240 @default.
- W3016188195 hasRelatedWork W3079443720 @default.
- W3016188195 isParatext "false" @default.
- W3016188195 isRetracted "false" @default.
- W3016188195 magId "3016188195" @default.
- W3016188195 workType "article" @default.