Matches in SemOpenAlex for { <https://semopenalex.org/work/W2922200935> ?p ?o ?g. }
Showing items 1 to 92 of
92
with 100 items per page.
- W2922200935 abstract "Long Short Term Memory Connectionist Temporal Classification (LSTM-CTC) based end-to-end models are widely used in speech recognition due to its simplicity in training and efficiency in decoding. In conventional LSTM-CTC based models, a bottleneck projection matrix maps the hidden feature vectors obtained from LSTM to softmax output layer. In this paper, we propose to use a high rank projection layer to replace the projection matrix. The output from the high rank projection layer is a weighted combination of vectors that are projected from the hidden feature vectors via different projection matrices and non-linear activation function. The high rank projection layer is able to improve the expressiveness of LSTM-CTC models. The experimental results show that on Wall Street Journal (WSJ) corpus and LibriSpeech data set, the proposed method achieves 4% - 6% relative word error rate (WER) reduction over the baseline CTC system. They outperform other published CTC based end-to-end (E2E) models under the condition that no external data or data augmentation is applied. Code has been made available at https://github.com/mobvoi/lstm_ctc." @default.
- W2922200935 created "2019-03-22" @default.
- W2922200935 creator A5030660132 @default.
- W2922200935 creator A5049457213 @default.
- W2922200935 creator A5085953379 @default.
- W2922200935 date "2019-05-01" @default.
- W2922200935 modified "2023-10-03" @default.
- W2922200935 title "End-to-end Speech Recognition Using a High Rank LSTM-CTC Based Model" @default.
- W2922200935 cites W2024490156 @default.
- W2922200935 cites W2127141656 @default.
- W2922200935 cites W2160815625 @default.
- W2922200935 cites W2327501763 @default.
- W2922200935 cites W2519224033 @default.
- W2922200935 cites W2526425061 @default.
- W2922200935 cites W2739883972 @default.
- W2922200935 cites W2746192915 @default.
- W2922200935 cites W2799800213 @default.
- W2922200935 cites W2962780374 @default.
- W2922200935 cites W2962962542 @default.
- W2922200935 cites W2963920996 @default.
- W2922200935 doi "https://doi.org/10.1109/icassp.2019.8683297" @default.
- W2922200935 hasPublicationYear "2019" @default.
- W2922200935 type Work @default.
- W2922200935 sameAs 2922200935 @default.
- W2922200935 citedByCount "8" @default.
- W2922200935 countsByYear W29222009352019 @default.
- W2922200935 countsByYear W29222009352021 @default.
- W2922200935 countsByYear W29222009352022 @default.
- W2922200935 countsByYear W29222009352023 @default.
- W2922200935 crossrefType "proceedings-article" @default.
- W2922200935 hasAuthorship W2922200935A5030660132 @default.
- W2922200935 hasAuthorship W2922200935A5049457213 @default.
- W2922200935 hasAuthorship W2922200935A5085953379 @default.
- W2922200935 hasBestOaLocation W29222009352 @default.
- W2922200935 hasConcept C108583219 @default.
- W2922200935 hasConcept C11413529 @default.
- W2922200935 hasConcept C114614502 @default.
- W2922200935 hasConcept C138885662 @default.
- W2922200935 hasConcept C149635348 @default.
- W2922200935 hasConcept C153180895 @default.
- W2922200935 hasConcept C154945302 @default.
- W2922200935 hasConcept C164226766 @default.
- W2922200935 hasConcept C188441871 @default.
- W2922200935 hasConcept C2776401178 @default.
- W2922200935 hasConcept C2780513914 @default.
- W2922200935 hasConcept C28490314 @default.
- W2922200935 hasConcept C33923547 @default.
- W2922200935 hasConcept C40969351 @default.
- W2922200935 hasConcept C41008148 @default.
- W2922200935 hasConcept C41895202 @default.
- W2922200935 hasConcept C50644808 @default.
- W2922200935 hasConcept C57273362 @default.
- W2922200935 hasConcept C57493831 @default.
- W2922200935 hasConcept C8521452 @default.
- W2922200935 hasConceptScore W2922200935C108583219 @default.
- W2922200935 hasConceptScore W2922200935C11413529 @default.
- W2922200935 hasConceptScore W2922200935C114614502 @default.
- W2922200935 hasConceptScore W2922200935C138885662 @default.
- W2922200935 hasConceptScore W2922200935C149635348 @default.
- W2922200935 hasConceptScore W2922200935C153180895 @default.
- W2922200935 hasConceptScore W2922200935C154945302 @default.
- W2922200935 hasConceptScore W2922200935C164226766 @default.
- W2922200935 hasConceptScore W2922200935C188441871 @default.
- W2922200935 hasConceptScore W2922200935C2776401178 @default.
- W2922200935 hasConceptScore W2922200935C2780513914 @default.
- W2922200935 hasConceptScore W2922200935C28490314 @default.
- W2922200935 hasConceptScore W2922200935C33923547 @default.
- W2922200935 hasConceptScore W2922200935C40969351 @default.
- W2922200935 hasConceptScore W2922200935C41008148 @default.
- W2922200935 hasConceptScore W2922200935C41895202 @default.
- W2922200935 hasConceptScore W2922200935C50644808 @default.
- W2922200935 hasConceptScore W2922200935C57273362 @default.
- W2922200935 hasConceptScore W2922200935C57493831 @default.
- W2922200935 hasConceptScore W2922200935C8521452 @default.
- W2922200935 hasLocation W29222009351 @default.
- W2922200935 hasLocation W29222009352 @default.
- W2922200935 hasOpenAccess W2922200935 @default.
- W2922200935 hasPrimaryLocation W29222009351 @default.
- W2922200935 hasRelatedWork W2100336732 @default.
- W2922200935 hasRelatedWork W2743258233 @default.
- W2922200935 hasRelatedWork W2782592381 @default.
- W2922200935 hasRelatedWork W2910692330 @default.
- W2922200935 hasRelatedWork W2921182884 @default.
- W2922200935 hasRelatedWork W2997969508 @default.
- W2922200935 hasRelatedWork W3202244193 @default.
- W2922200935 hasRelatedWork W3208883981 @default.
- W2922200935 hasRelatedWork W4307834408 @default.
- W2922200935 hasRelatedWork W4320925816 @default.
- W2922200935 isParatext "false" @default.
- W2922200935 isRetracted "false" @default.
- W2922200935 magId "2922200935" @default.
- W2922200935 workType "article" @default.