Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385194792> ?p ?o ?g. }
- W4385194792 endingPage "936" @default.
- W4385194792 startingPage "932" @default.
- W4385194792 abstract "In recent years, voice pathology detection (VPD) has received considerable attention because of the increasing risk of voice problems. Several methods, such as support vector machine and convolutional neural network-based models, achieve good VPD performance. To further improve the performance, we use a self-supervised pretrained model as feature representation instead of explicit speech features. When the pretrained model is fine-tuned for VPD, an overfitting problem occurs due to a domain shift from conversation speech to the VPD task. To mitigate this problem, we propose an adversarial task adaptive pretraining (A-TAPT) approach by incorporating adversarial regularization during the continual learning process. Experiments on VPD using the Saarbrucken Voice Database show that the proposed A-TAPT improves the unweighted average recall (UAR) by an absolute increase of 12.36% and 15.38% compared with SVM and ResNet50, respectively. It is also shown that the proposed A-TAPT achieves a UAR that is 2.77% higher than that of conventional TAPT learning." @default.
- W4385194792 created "2023-07-25" @default.
- W4385194792 creator A5052959289 @default.
- W4385194792 creator A5080463589 @default.
- W4385194792 creator A5084586005 @default.
- W4385194792 creator A5085604883 @default.
- W4385194792 date "2023-01-01" @default.
- W4385194792 modified "2023-10-02" @default.
- W4385194792 title "Adversarial Continual Learning to Transfer Self-Supervised Speech Representations for Voice Pathology Detection" @default.
- W4385194792 cites W1494198834 @default.
- W4385194792 cites W2218075793 @default.
- W4385194792 cites W2339666751 @default.
- W4385194792 cites W2385755489 @default.
- W4385194792 cites W2581209700 @default.
- W4385194792 cites W2609527102 @default.
- W4385194792 cites W2798991696 @default.
- W4385194792 cites W2807317037 @default.
- W4385194792 cites W2963026768 @default.
- W4385194792 cites W2963855133 @default.
- W4385194792 cites W3014146919 @default.
- W4385194792 cites W3031948080 @default.
- W4385194792 cites W3034238904 @default.
- W4385194792 cites W3035204084 @default.
- W4385194792 cites W3095385732 @default.
- W4385194792 cites W3196473068 @default.
- W4385194792 cites W3196525293 @default.
- W4385194792 cites W3198275944 @default.
- W4385194792 cites W3198712976 @default.
- W4385194792 cites W3204696009 @default.
- W4385194792 cites W3204774355 @default.
- W4385194792 cites W3207558756 @default.
- W4385194792 cites W3209984917 @default.
- W4385194792 cites W4226380987 @default.
- W4385194792 cites W4297841902 @default.
- W4385194792 cites W4308599792 @default.
- W4385194792 cites W4312259157 @default.
- W4385194792 cites W4319878709 @default.
- W4385194792 cites W4372338343 @default.
- W4385194792 cites W4375869379 @default.
- W4385194792 doi "https://doi.org/10.1109/lsp.2023.3298532" @default.
- W4385194792 hasPublicationYear "2023" @default.
- W4385194792 type Work @default.
- W4385194792 citedByCount "0" @default.
- W4385194792 crossrefType "journal-article" @default.
- W4385194792 hasAuthorship W4385194792A5052959289 @default.
- W4385194792 hasAuthorship W4385194792A5080463589 @default.
- W4385194792 hasAuthorship W4385194792A5084586005 @default.
- W4385194792 hasAuthorship W4385194792A5085604883 @default.
- W4385194792 hasConcept C108583219 @default.
- W4385194792 hasConcept C119857082 @default.
- W4385194792 hasConcept C12267149 @default.
- W4385194792 hasConcept C138885662 @default.
- W4385194792 hasConcept C150899416 @default.
- W4385194792 hasConcept C153180895 @default.
- W4385194792 hasConcept C154945302 @default.
- W4385194792 hasConcept C162324750 @default.
- W4385194792 hasConcept C187736073 @default.
- W4385194792 hasConcept C204201278 @default.
- W4385194792 hasConcept C22019652 @default.
- W4385194792 hasConcept C2776135515 @default.
- W4385194792 hasConcept C2777200299 @default.
- W4385194792 hasConcept C2780451532 @default.
- W4385194792 hasConcept C28006648 @default.
- W4385194792 hasConcept C28490314 @default.
- W4385194792 hasConcept C37736160 @default.
- W4385194792 hasConcept C41008148 @default.
- W4385194792 hasConcept C41895202 @default.
- W4385194792 hasConcept C50644808 @default.
- W4385194792 hasConcept C59404180 @default.
- W4385194792 hasConcept C61328038 @default.
- W4385194792 hasConcept C81363708 @default.
- W4385194792 hasConceptScore W4385194792C108583219 @default.
- W4385194792 hasConceptScore W4385194792C119857082 @default.
- W4385194792 hasConceptScore W4385194792C12267149 @default.
- W4385194792 hasConceptScore W4385194792C138885662 @default.
- W4385194792 hasConceptScore W4385194792C150899416 @default.
- W4385194792 hasConceptScore W4385194792C153180895 @default.
- W4385194792 hasConceptScore W4385194792C154945302 @default.
- W4385194792 hasConceptScore W4385194792C162324750 @default.
- W4385194792 hasConceptScore W4385194792C187736073 @default.
- W4385194792 hasConceptScore W4385194792C204201278 @default.
- W4385194792 hasConceptScore W4385194792C22019652 @default.
- W4385194792 hasConceptScore W4385194792C2776135515 @default.
- W4385194792 hasConceptScore W4385194792C2777200299 @default.
- W4385194792 hasConceptScore W4385194792C2780451532 @default.
- W4385194792 hasConceptScore W4385194792C28006648 @default.
- W4385194792 hasConceptScore W4385194792C28490314 @default.
- W4385194792 hasConceptScore W4385194792C37736160 @default.
- W4385194792 hasConceptScore W4385194792C41008148 @default.
- W4385194792 hasConceptScore W4385194792C41895202 @default.
- W4385194792 hasConceptScore W4385194792C50644808 @default.
- W4385194792 hasConceptScore W4385194792C59404180 @default.
- W4385194792 hasConceptScore W4385194792C61328038 @default.
- W4385194792 hasConceptScore W4385194792C81363708 @default.
- W4385194792 hasLocation W43851947921 @default.
- W4385194792 hasOpenAccess W4385194792 @default.
- W4385194792 hasPrimaryLocation W43851947921 @default.
- W4385194792 hasRelatedWork W2767651786 @default.