FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2547335510> ?p ?o ?g. }

• W2547335510 abstract "As has been extensively shown, acoustic features for speech recognition can be nurtured from training data using neural networks (DNN) with multiple hidden layers. Although a large body of research has shown these learned features are superior to standard front- ends, this superiority is usually demonstrated when the data used to learn the features is very similar to the data used to test recognition performance. However, realistic environments cover many unanticipated types of novel inputs including noise, channel distortion, reverberation, accented speech, speaking rate variation, overlapped speech, etc. A quantitative analysis using bootstrap sampling shows that these trained features are easily specialized to training data and corrupted in mismatched scenarios. Gabor filtered spectrograms, on the other hand, are generated from spectro-temporal filters to model natural human auditory processing, which can be instrumental in improving generalization to unanticipated deviations from what was seen in training. In this thesis, I used Gabor filtering as feature processing or a convolutional kernel in neural networks where the former used filter outputs as DNN inputs while the latter used filter coefficients and structures to initialize a convolutional neural network (CNN). Experiments show that the proposed features perform better than other noise-robust features that I have tried on several noisy corpora. In addition, I demonstrate that inclusion of Gabor filters with lower or higher temporal modulations could be used to correlate better with human perception of slow or rapid speech. Finally, I report on the analysis of human cortical signals to demonstrate the relative robustness of these signals to the mixed signal phenomenon in contrast to a DNN-based ASR system. With a number of example tasks in the thesis, I conclude that designed feature is useful for greater robustness than just relying on DNN or CNN." @default.
• W2547335510 created "2016-11-11" @default.
• W2547335510 creator A5001306222 @default.
• W2547335510 date "2016-01-01" @default.
• W2547335510 modified "2023-09-27" @default.
• W2547335510 title "Feature Design for Robust Speech Recognition: Nurture and Nature" @default.
• W2547335510 cites W1492775261 @default.
• W2547335510 cites W1495061682 @default.
• W2547335510 cites W1524333225 @default.
• W2547335510 cites W1542280630 @default.
• W2547335510 cites W1553004968 @default.
• W2547335510 cites W1576884575 @default.
• W2547335510 cites W1591607137 @default.
• W2547335510 cites W1627087495 @default.
• W2547335510 cites W1665214252 @default.
• W2547335510 cites W1902027874 @default.
• W2547335510 cites W1974387177 @default.
• W2547335510 cites W1978741356 @default.
• W2547335510 cites W1990005915 @default.
• W2547335510 cites W1991181922 @default.
• W2547335510 cites W1993882792 @default.
• W2547335510 cites W1997320786 @default.
• W2547335510 cites W2008505918 @default.
• W2547335510 cites W2018742520 @default.
• W2547335510 cites W2040870580 @default.
• W2547335510 cites W2046058153 @default.
• W2547335510 cites W2048608953 @default.
• W2547335510 cites W2054883573 @default.
• W2547335510 cites W2055910046 @default.
• W2547335510 cites W2074250290 @default.
• W2547335510 cites W2079207700 @default.
• W2547335510 cites W2082183045 @default.
• W2547335510 cites W2088489891 @default.
• W2547335510 cites W2100495367 @default.
• W2547335510 cites W2112739286 @default.
• W2547335510 cites W2112796928 @default.
• W2547335510 cites W2113427327 @default.
• W2547335510 cites W2116742884 @default.
• W2547335510 cites W2117671523 @default.
• W2547335510 cites W2118980133 @default.
• W2547335510 cites W2119673740 @default.
• W2547335510 cites W2124181495 @default.
• W2547335510 cites W2130426352 @default.
• W2547335510 cites W2138584058 @default.
• W2547335510 cites W2140539590 @default.
• W2547335510 cites W2140595311 @default.
• W2547335510 cites W2140979961 @default.
• W2547335510 cites W2141778357 @default.
• W2547335510 cites W2142339048 @default.
• W2547335510 cites W2143612262 @default.
• W2547335510 cites W2151484683 @default.
• W2547335510 cites W2154649497 @default.
• W2547335510 cites W2156615793 @default.
• W2547335510 cites W2159948109 @default.
• W2547335510 cites W2160783467 @default.
• W2547335510 cites W2161224286 @default.
• W2547335510 cites W2163929346 @default.
• W2547335510 cites W2165712214 @default.
• W2547335510 cites W2167763959 @default.
• W2547335510 cites W2168103112 @default.
• W2547335510 cites W2171928131 @default.
• W2547335510 cites W2171965202 @default.
• W2547335510 cites W2172097686 @default.
• W2547335510 cites W2188183693 @default.
• W2547335510 cites W2322002063 @default.
• W2547335510 cites W2394932179 @default.
• W2547335510 cites W2400997536 @default.
• W2547335510 cites W2404548287 @default.
• W2547335510 cites W2407835985 @default.
• W2547335510 cites W2887979278 @default.
• W2547335510 cites W3099514962 @default.
• W2547335510 cites W42203131 @default.
• W2547335510 hasPublicationYear "2016" @default.
• W2547335510 type Work @default.
• W2547335510 sameAs 2547335510 @default.
• W2547335510 citedByCount "0" @default.
• W2547335510 crossrefType "journal-article" @default.
• W2547335510 hasAuthorship W2547335510A5001306222 @default.
• W2547335510 hasConcept C104317684 @default.
• W2547335510 hasConcept C121332964 @default.
• W2547335510 hasConcept C138885662 @default.
• W2547335510 hasConcept C153180895 @default.
• W2547335510 hasConcept C154945302 @default.
• W2547335510 hasConcept C185592680 @default.
• W2547335510 hasConcept C24890656 @default.
• W2547335510 hasConcept C2776401178 @default.
• W2547335510 hasConcept C28490314 @default.
• W2547335510 hasConcept C41008148 @default.
• W2547335510 hasConcept C41895202 @default.
• W2547335510 hasConcept C45273575 @default.
• W2547335510 hasConcept C50644808 @default.
• W2547335510 hasConcept C55493867 @default.
• W2547335510 hasConcept C63479239 @default.
• W2547335510 hasConcept C81363708 @default.
• W2547335510 hasConcept C95851461 @default.
• W2547335510 hasConceptScore W2547335510C104317684 @default.
• W2547335510 hasConceptScore W2547335510C121332964 @default.
• W2547335510 hasConceptScore W2547335510C138885662 @default.
• W2547335510 hasConceptScore W2547335510C153180895 @default.
• W2547335510 hasConceptScore W2547335510C154945302 @default.

• next