Matches in SemOpenAlex for { <https://semopenalex.org/work/W2752168051> ?p ?o ?g. }
- W2752168051 abstract "Neural models have become ubiquitous in automatic speech recognition systems. While neural networks are typically used as acoustic models in more complex systems, recent studies have explored end-to-end speech recognition systems based on neural networks, which can be trained to directly predict text from input acoustic features. Although such systems are conceptually elegant and simpler than traditional systems, it is less obvious how to interpret the trained models. In this work, we analyze the speech representations learned by a deep end-to-end model that is based on convolutional and recurrent layers, and trained with a connectionist temporal classification (CTC) loss. We use a pre-trained model to generate frame-level features which are given to a classifier that is trained on frame classification into phones. We evaluate representations from different layers of the deep model and compare their quality for predicting phone labels. Our experiments shed light on important aspects of the end-to-end model such as layer depth, model complexity, and other design choices." @default.
- W2752168051 created "2017-09-15" @default.
- W2752168051 creator A5039081803 @default.
- W2752168051 creator A5051184573 @default.
- W2752168051 date "2017-09-13" @default.
- W2752168051 modified "2023-09-27" @default.
- W2752168051 title "Analyzing Hidden Representations in End-to-End Automatic Speech Recognition Systems" @default.
- W2752168051 cites W1494198834 @default.
- W2752168051 cites W1522301498 @default.
- W2752168051 cites W1586532344 @default.
- W2752168051 cites W1600744878 @default.
- W2752168051 cites W1736701665 @default.
- W2752168051 cites W2064675550 @default.
- W2752168051 cites W2077804127 @default.
- W2752168051 cites W2102113734 @default.
- W2752168051 cites W2127141656 @default.
- W2752168051 cites W2158373110 @default.
- W2752168051 cites W2172097686 @default.
- W2752168051 cites W2187089797 @default.
- W2752168051 cites W2295676751 @default.
- W2752168051 cites W2327501763 @default.
- W2752168051 cites W2531381952 @default.
- W2752168051 cites W2545177271 @default.
- W2752168051 cites W2563574619 @default.
- W2752168051 cites W2580178245 @default.
- W2752168051 cites W2607361225 @default.
- W2752168051 cites W2949117887 @default.
- W2752168051 cites W2962826786 @default.
- W2752168051 cites W2962949994 @default.
- W2752168051 cites W2963583362 @default.
- W2752168051 cites W2964060510 @default.
- W2752168051 cites W3105148948 @default.
- W2752168051 hasPublicationYear "2017" @default.
- W2752168051 type Work @default.
- W2752168051 sameAs 2752168051 @default.
- W2752168051 citedByCount "1" @default.
- W2752168051 countsByYear W27521680512018 @default.
- W2752168051 crossrefType "posted-content" @default.
- W2752168051 hasAuthorship W2752168051A5039081803 @default.
- W2752168051 hasAuthorship W2752168051A5051184573 @default.
- W2752168051 hasConcept C108583219 @default.
- W2752168051 hasConcept C119857082 @default.
- W2752168051 hasConcept C126042441 @default.
- W2752168051 hasConcept C153180895 @default.
- W2752168051 hasConcept C154945302 @default.
- W2752168051 hasConcept C155635449 @default.
- W2752168051 hasConcept C23224414 @default.
- W2752168051 hasConcept C28490314 @default.
- W2752168051 hasConcept C2984842247 @default.
- W2752168051 hasConcept C41008148 @default.
- W2752168051 hasConcept C50644808 @default.
- W2752168051 hasConcept C61328038 @default.
- W2752168051 hasConcept C74296488 @default.
- W2752168051 hasConcept C76155785 @default.
- W2752168051 hasConcept C81363708 @default.
- W2752168051 hasConcept C8521452 @default.
- W2752168051 hasConcept C95623464 @default.
- W2752168051 hasConceptScore W2752168051C108583219 @default.
- W2752168051 hasConceptScore W2752168051C119857082 @default.
- W2752168051 hasConceptScore W2752168051C126042441 @default.
- W2752168051 hasConceptScore W2752168051C153180895 @default.
- W2752168051 hasConceptScore W2752168051C154945302 @default.
- W2752168051 hasConceptScore W2752168051C155635449 @default.
- W2752168051 hasConceptScore W2752168051C23224414 @default.
- W2752168051 hasConceptScore W2752168051C28490314 @default.
- W2752168051 hasConceptScore W2752168051C2984842247 @default.
- W2752168051 hasConceptScore W2752168051C41008148 @default.
- W2752168051 hasConceptScore W2752168051C50644808 @default.
- W2752168051 hasConceptScore W2752168051C61328038 @default.
- W2752168051 hasConceptScore W2752168051C74296488 @default.
- W2752168051 hasConceptScore W2752168051C76155785 @default.
- W2752168051 hasConceptScore W2752168051C81363708 @default.
- W2752168051 hasConceptScore W2752168051C8521452 @default.
- W2752168051 hasConceptScore W2752168051C95623464 @default.
- W2752168051 hasLocation W27521680511 @default.
- W2752168051 hasOpenAccess W2752168051 @default.
- W2752168051 hasPrimaryLocation W27521680511 @default.
- W2752168051 hasRelatedWork W2138104661 @default.
- W2752168051 hasRelatedWork W2513602745 @default.
- W2752168051 hasRelatedWork W2610961739 @default.
- W2752168051 hasRelatedWork W2625426967 @default.
- W2752168051 hasRelatedWork W2739693977 @default.
- W2752168051 hasRelatedWork W2799034895 @default.
- W2752168051 hasRelatedWork W2803039862 @default.
- W2752168051 hasRelatedWork W2895358653 @default.
- W2752168051 hasRelatedWork W2899135636 @default.
- W2752168051 hasRelatedWork W2900836049 @default.
- W2752168051 hasRelatedWork W2962813140 @default.
- W2752168051 hasRelatedWork W2973037561 @default.
- W2752168051 hasRelatedWork W3007253895 @default.
- W2752168051 hasRelatedWork W3096048508 @default.
- W2752168051 hasRelatedWork W3154658447 @default.
- W2752168051 hasRelatedWork W3165160356 @default.
- W2752168051 hasRelatedWork W3166052096 @default.
- W2752168051 hasRelatedWork W385738233 @default.
- W2752168051 hasRelatedWork W392164126 @default.
- W2752168051 hasRelatedWork W80990663 @default.
- W2752168051 isParatext "false" @default.
- W2752168051 isRetracted "false" @default.
- W2752168051 magId "2752168051" @default.