Matches in SemOpenAlex for { <https://semopenalex.org/work/W3105148948> ?p ?o ?g. }
- W3105148948 abstract "We study the representation and encoding of phonemes in a recurrent neural network model of grounded speech. We use a model which processes images and their spoken descriptions, and projects the visual and auditory representations into the same semantic space. We perform a number of analyses on how information about individual phonemes is encoded in the MFCC features extracted from the speech signal, and the activations of the layers of the model. Via experiments with phoneme decoding and phoneme discrimination we show that phoneme representations are most salient in the lower layers of the model, where low-level signals are processed at a fine-grained level, although a large amount of phonological information is retain at the top recurrent layer. We further find out that the attention mechanism following the top recurrent layer significantly attenuates encoding of phonology and makes the utterance embeddings much more invariant to synonymy. Moreover, a hierarchical clustering of phoneme representations learned by the network shows an organizational structure of phonemes similar to those proposed in linguistics." @default.
- W3105148948 created "2020-11-23" @default.
- W3105148948 creator A5022698890 @default.
- W3105148948 creator A5047382001 @default.
- W3105148948 creator A5055265528 @default.
- W3105148948 date "2017-01-01" @default.
- W3105148948 modified "2023-10-12" @default.
- W3105148948 title "Encoding of phonology in a recurrent neural model of grounded speech" @default.
- W3105148948 cites W1524333225 @default.
- W3105148948 cites W1606347560 @default.
- W3105148948 cites W1686810756 @default.
- W3105148948 cites W1861492603 @default.
- W3105148948 cites W1988126949 @default.
- W3105148948 cites W1992570774 @default.
- W3105148948 cites W2011238950 @default.
- W3105148948 cites W2016381774 @default.
- W3105148948 cites W2050789388 @default.
- W3105148948 cites W2068247585 @default.
- W3105148948 cites W2081580037 @default.
- W3105148948 cites W2087946919 @default.
- W3105148948 cites W2096103549 @default.
- W3105148948 cites W2107917162 @default.
- W3105148948 cites W2114338548 @default.
- W3105148948 cites W2115646488 @default.
- W3105148948 cites W2123815913 @default.
- W3105148948 cites W2127846433 @default.
- W3105148948 cites W2137010615 @default.
- W3105148948 cites W2143017621 @default.
- W3105148948 cites W2151311453 @default.
- W3105148948 cites W2172097686 @default.
- W3105148948 cites W2173628018 @default.
- W3105148948 cites W2292919134 @default.
- W3105148948 cites W2343494948 @default.
- W3105148948 cites W2395899413 @default.
- W3105148948 cites W2473934411 @default.
- W3105148948 cites W2515741950 @default.
- W3105148948 cites W2531381952 @default.
- W3105148948 cites W2549835527 @default.
- W3105148948 cites W2556930864 @default.
- W3105148948 cites W2562979205 @default.
- W3105148948 cites W2580178245 @default.
- W3105148948 cites W2582895110 @default.
- W3105148948 cites W2589031409 @default.
- W3105148948 cites W2952020226 @default.
- W3105148948 cites W2962776659 @default.
- W3105148948 cites W2962835968 @default.
- W3105148948 cites W2962862718 @default.
- W3105148948 cites W2964159778 @default.
- W3105148948 cites W3213502289 @default.
- W3105148948 doi "https://doi.org/10.18653/v1/k17-1037" @default.
- W3105148948 hasPublicationYear "2017" @default.
- W3105148948 type Work @default.
- W3105148948 sameAs 3105148948 @default.
- W3105148948 citedByCount "52" @default.
- W3105148948 countsByYear W31051489482017 @default.
- W3105148948 countsByYear W31051489482018 @default.
- W3105148948 countsByYear W31051489482019 @default.
- W3105148948 countsByYear W31051489482020 @default.
- W3105148948 countsByYear W31051489482021 @default.
- W3105148948 countsByYear W31051489482022 @default.
- W3105148948 countsByYear W31051489482023 @default.
- W3105148948 crossrefType "proceedings-article" @default.
- W3105148948 hasAuthorship W3105148948A5022698890 @default.
- W3105148948 hasAuthorship W3105148948A5047382001 @default.
- W3105148948 hasAuthorship W3105148948A5055265528 @default.
- W3105148948 hasBestOaLocation W31051489481 @default.
- W3105148948 hasConcept C125411270 @default.
- W3105148948 hasConcept C138885662 @default.
- W3105148948 hasConcept C147168706 @default.
- W3105148948 hasConcept C148934300 @default.
- W3105148948 hasConcept C14999030 @default.
- W3105148948 hasConcept C154945302 @default.
- W3105148948 hasConcept C17744445 @default.
- W3105148948 hasConcept C199539241 @default.
- W3105148948 hasConcept C204321447 @default.
- W3105148948 hasConcept C2775852435 @default.
- W3105148948 hasConcept C2776359362 @default.
- W3105148948 hasConcept C2780719617 @default.
- W3105148948 hasConcept C28490314 @default.
- W3105148948 hasConcept C41008148 @default.
- W3105148948 hasConcept C41895202 @default.
- W3105148948 hasConcept C50644808 @default.
- W3105148948 hasConcept C57273362 @default.
- W3105148948 hasConcept C76155785 @default.
- W3105148948 hasConcept C91863865 @default.
- W3105148948 hasConcept C94625758 @default.
- W3105148948 hasConceptScore W3105148948C125411270 @default.
- W3105148948 hasConceptScore W3105148948C138885662 @default.
- W3105148948 hasConceptScore W3105148948C147168706 @default.
- W3105148948 hasConceptScore W3105148948C148934300 @default.
- W3105148948 hasConceptScore W3105148948C14999030 @default.
- W3105148948 hasConceptScore W3105148948C154945302 @default.
- W3105148948 hasConceptScore W3105148948C17744445 @default.
- W3105148948 hasConceptScore W3105148948C199539241 @default.
- W3105148948 hasConceptScore W3105148948C204321447 @default.
- W3105148948 hasConceptScore W3105148948C2775852435 @default.
- W3105148948 hasConceptScore W3105148948C2776359362 @default.
- W3105148948 hasConceptScore W3105148948C2780719617 @default.
- W3105148948 hasConceptScore W3105148948C28490314 @default.
- W3105148948 hasConceptScore W3105148948C41008148 @default.