Matches in SemOpenAlex for { <https://semopenalex.org/work/W3097414768> ?p ?o ?g. }
Showing items 1 to 64 of
64
with 100 items per page.
- W3097414768 abstract "Speech is one of the most effective means of communication and is full of information that helps the transmission of utterer's thoughts. However, mainly due to the cumbersome processing of acoustic features, phoneme or word posterior probability has frequently been discarded in understanding the natural language. Thus, some recent spoken language understanding (SLU) modules have utilized end-to-end structures that preserve the uncertainty information. This further reduces the propagation of speech recognition error and guarantees computational efficiency. We claim that in this process, the speech comprehension can benefit from the inference of massive pre-trained language models (LMs). We transfer the knowledge from a concrete Transformer-based text LM to an SLU module which can face a data shortage, based on recent cross-modal distillation methodologies. We demonstrate the validity of our proposal upon the performance on Fluent Speech Command, an English SLU benchmark. Thereby, we experimentally verify our hypothesis that the knowledge could be shared from the top layer of the LM to a fully speech-based module, in which the abstracted speech is expected to meet the semantic representation." @default.
- W3097414768 created "2020-11-09" @default.
- W3097414768 creator A5005340561 @default.
- W3097414768 creator A5016783920 @default.
- W3097414768 creator A5044323454 @default.
- W3097414768 creator A5051356511 @default.
- W3097414768 date "2020-10-25" @default.
- W3097414768 modified "2023-10-16" @default.
- W3097414768 title "Speech to Text Adaptation: Towards an Efficient Cross-Modal Distillation" @default.
- W3097414768 doi "https://doi.org/10.21437/interspeech.2020-1246" @default.
- W3097414768 hasPublicationYear "2020" @default.
- W3097414768 type Work @default.
- W3097414768 sameAs 3097414768 @default.
- W3097414768 citedByCount "12" @default.
- W3097414768 countsByYear W30974147682020 @default.
- W3097414768 countsByYear W30974147682021 @default.
- W3097414768 countsByYear W30974147682022 @default.
- W3097414768 countsByYear W30974147682023 @default.
- W3097414768 crossrefType "proceedings-article" @default.
- W3097414768 hasAuthorship W3097414768A5005340561 @default.
- W3097414768 hasAuthorship W3097414768A5016783920 @default.
- W3097414768 hasAuthorship W3097414768A5044323454 @default.
- W3097414768 hasAuthorship W3097414768A5051356511 @default.
- W3097414768 hasBestOaLocation W30974147682 @default.
- W3097414768 hasConcept C139807058 @default.
- W3097414768 hasConcept C15744967 @default.
- W3097414768 hasConcept C169760540 @default.
- W3097414768 hasConcept C178790620 @default.
- W3097414768 hasConcept C185592680 @default.
- W3097414768 hasConcept C188027245 @default.
- W3097414768 hasConcept C204030448 @default.
- W3097414768 hasConcept C204321447 @default.
- W3097414768 hasConcept C28490314 @default.
- W3097414768 hasConcept C41008148 @default.
- W3097414768 hasConcept C71139939 @default.
- W3097414768 hasConceptScore W3097414768C139807058 @default.
- W3097414768 hasConceptScore W3097414768C15744967 @default.
- W3097414768 hasConceptScore W3097414768C169760540 @default.
- W3097414768 hasConceptScore W3097414768C178790620 @default.
- W3097414768 hasConceptScore W3097414768C185592680 @default.
- W3097414768 hasConceptScore W3097414768C188027245 @default.
- W3097414768 hasConceptScore W3097414768C204030448 @default.
- W3097414768 hasConceptScore W3097414768C204321447 @default.
- W3097414768 hasConceptScore W3097414768C28490314 @default.
- W3097414768 hasConceptScore W3097414768C41008148 @default.
- W3097414768 hasConceptScore W3097414768C71139939 @default.
- W3097414768 hasLocation W30974147681 @default.
- W3097414768 hasLocation W30974147682 @default.
- W3097414768 hasOpenAccess W3097414768 @default.
- W3097414768 hasPrimaryLocation W30974147681 @default.
- W3097414768 hasRelatedWork W1543657218 @default.
- W3097414768 hasRelatedWork W1552159754 @default.
- W3097414768 hasRelatedWork W1571518467 @default.
- W3097414768 hasRelatedWork W189527659 @default.
- W3097414768 hasRelatedWork W2020291234 @default.
- W3097414768 hasRelatedWork W2135121413 @default.
- W3097414768 hasRelatedWork W2368651715 @default.
- W3097414768 hasRelatedWork W2611614995 @default.
- W3097414768 hasRelatedWork W2789919619 @default.
- W3097414768 hasRelatedWork W87991986 @default.
- W3097414768 isParatext "false" @default.
- W3097414768 isRetracted "false" @default.
- W3097414768 magId "3097414768" @default.
- W3097414768 workType "article" @default.