FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3103003256> ?p ?o ?g. }

• W3103003256 endingPage "117" @default.
• W3103003256 startingPage "106" @default.
• W3103003256 abstract "A new cross-entropy-guided measure (CEGM) is proposed to indirectly assess accuracies of automatic speech recognition (ASR) of degraded speech with a speech enhancement front-end and without directly performing ASR experiments. The proposed CEGM is calculated in three steps, namely: (1) a low-level representations via feature extraction, (2) a high-level nonlinear mapping using an acoustic model, and (3) a final CEGM calculation between the high-level representations of clean and enhanced speech. Specifically, state posterior probabilities from outputs of conventional hybrid acoustic model of the target ASR system are adopted as the high-level representations and a cross-entropy criterion is used to calculate the CEGM. Due to CEGM's differentiability, it can also be used to replace the conventional minimum mean squared error (MMSE) criterion as an objective function for deep neural network (DNN)-based speech enhancement. Therefore, the front-end enhancement model can be optimized towards improving the accuracies of the back-end ASR system. Experiments on single-channel CHiME-4 Challenge show that CEGM yields consistently the highest correlations with word error rate (WER) which is often costly to calculate, and achieves the most accurate assessment of ASR performance when compared to the perceptual evaluation metrics commonly used for assessing speech enhancement performance. Furthermore, CEGM-optimized speech enhancement could effectively reduce the WER on the CHiME-4 real test set when compared to unprocessed noisy speech and enhanced speech obtained with MMSE-optimized enhancement for ASR systems with fixed multi-condition acoustic models in various deep architectures." @default.
• W3103003256 created "2020-11-23" @default.
• W3103003256 creator A5039181878 @default.
• W3103003256 creator A5051261170 @default.
• W3103003256 creator A5066595711 @default.
• W3103003256 creator A5066868860 @default.
• W3103003256 date "2021-01-01" @default.
• W3103003256 modified "2023-09-23" @default.
• W3103003256 title "A Cross-Entropy-Guided Measure (CEGM) for Assessing Speech Recognition Performance and Optimizing DNN-Based Speech Enhancement" @default.
• W3103003256 cites W1494198834 @default.
• W3103003256 cites W1523493493 @default.
• W3103003256 cites W1598508708 @default.
• W3103003256 cites W1897240248 @default.
• W3103003256 cites W1901129140 @default.
• W3103003256 cites W1975163393 @default.
• W3103003256 cites W1988961123 @default.
• W3103003256 cites W1995536493 @default.
• W3103003256 cites W1999841894 @default.
• W3103003256 cites W2031692264 @default.
• W3103003256 cites W2039716246 @default.
• W3103003256 cites W2044893557 @default.
• W3103003256 cites W2045956438 @default.
• W3103003256 cites W2062164080 @default.
• W3103003256 cites W2069681747 @default.
• W3103003256 cites W2103264249 @default.
• W3103003256 cites W2113556376 @default.
• W3103003256 cites W2115415333 @default.
• W3103003256 cites W2126942983 @default.
• W3103003256 cites W2127851351 @default.
• W3103003256 cites W2128653836 @default.
• W3103003256 cites W2137988753 @default.
• W3103003256 cites W2141998673 @default.
• W3103003256 cites W2146324387 @default.
• W3103003256 cites W2153038597 @default.
• W3103003256 cites W2158501878 @default.
• W3103003256 cites W2160815625 @default.
• W3103003256 cites W2168379380 @default.
• W3103003256 cites W2296167893 @default.
• W3103003256 cites W2398042854 @default.
• W3103003256 cites W2400339399 @default.
• W3103003256 cites W2491975496 @default.
• W3103003256 cites W2517616541 @default.
• W3103003256 cites W2559260703 @default.
• W3103003256 cites W2560747691 @default.
• W3103003256 cites W2584390378 @default.
• W3103003256 cites W2613006358 @default.
• W3103003256 cites W2653598178 @default.
• W3103003256 cites W2711861986 @default.
• W3103003256 cites W2735006420 @default.
• W3103003256 cites W2889500840 @default.
• W3103003256 cites W2899829828 @default.
• W3103003256 cites W2937998856 @default.
• W3103003256 cites W2962843322 @default.
• W3103003256 cites W2962892438 @default.
• W3103003256 cites W2963045393 @default.
• W3103003256 cites W2963458655 @default.
• W3103003256 cites W2963828919 @default.
• W3103003256 cites W2969917057 @default.
• W3103003256 cites W2972320711 @default.
• W3103003256 cites W2972686263 @default.
• W3103003256 cites W2972990825 @default.
• W3103003256 cites W2973050374 @default.
• W3103003256 cites W3147539069 @default.
• W3103003256 doi "https://doi.org/10.1109/taslp.2020.3036783" @default.
• W3103003256 hasPublicationYear "2021" @default.
• W3103003256 type Work @default.
• W3103003256 sameAs 3103003256 @default.
• W3103003256 citedByCount "23" @default.
• W3103003256 countsByYear W31030032562021 @default.
• W3103003256 countsByYear W31030032562022 @default.
• W3103003256 countsByYear W31030032562023 @default.
• W3103003256 crossrefType "journal-article" @default.
• W3103003256 hasAuthorship W3103003256A5039181878 @default.
• W3103003256 hasAuthorship W3103003256A5051261170 @default.
• W3103003256 hasAuthorship W3103003256A5066595711 @default.
• W3103003256 hasAuthorship W3103003256A5066868860 @default.
• W3103003256 hasConcept C124101348 @default.
• W3103003256 hasConcept C154945302 @default.
• W3103003256 hasConcept C2780009758 @default.
• W3103003256 hasConcept C28490314 @default.
• W3103003256 hasConcept C41008148 @default.
• W3103003256 hasConceptScore W3103003256C124101348 @default.
• W3103003256 hasConceptScore W3103003256C154945302 @default.
• W3103003256 hasConceptScore W3103003256C2780009758 @default.
• W3103003256 hasConceptScore W3103003256C28490314 @default.
• W3103003256 hasConceptScore W3103003256C41008148 @default.
• W3103003256 hasFunder F4320321001 @default.
• W3103003256 hasLocation W31030032561 @default.
• W3103003256 hasOpenAccess W3103003256 @default.
• W3103003256 hasPrimaryLocation W31030032561 @default.
• W3103003256 hasRelatedWork W2312116756 @default.
• W3103003256 hasRelatedWork W2368779261 @default.
• W3103003256 hasRelatedWork W2381980384 @default.
• W3103003256 hasRelatedWork W2422371469 @default.
• W3103003256 hasRelatedWork W2778699561 @default.
• W3103003256 hasRelatedWork W2794438528 @default.
• W3103003256 hasRelatedWork W2893763841 @default.
• W3103003256 hasRelatedWork W2995996972 @default.

• next