Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387609344> ?p ?o ?g. }
Showing items 1 to 68 of
68
with 100 items per page.
- W4387609344 endingPage "12" @default.
- W4387609344 startingPage "1" @default.
- W4387609344 abstract "Given a reference speech clip from the target speaker, Target Speaker Extraction (TSE) is a challenging task that involves extracting the signal of the target speaker from a multi-speaker environment. TSE networks typically comprise a main network and an auxiliary network. The former utilizes the obtained target speaker embedding to generate an appropriate mask for isolating the signal of the target speaker from those of other speakers, while the latter aims to learn deep discriminative embeddings from the signal of the target speaker. However, the TSE networks often face performance degradation when dealing with unseen speakers or speeches with short references. In this paper, we propose a novel approach that leverages contrastive learning in the auxiliary network to obtain better representations of unseen speakers or speeches with short references. Specifically, we employ contrastive learning to bridge the gap between short and long speech features. In this case, the auxiliary network with the input of a short speech generates feature embeddings that are as rich as those obtained from a long speech. Therefore, improving the recognition of unseen speakers or short speeches. Moreover, we introduce an attention-based fusion method that integrates the speaker embedding into the main network in an adaptive manner for enhancing mask generation. Experimental results demonstrate the effectiveness of our proposed method in improving the performance of TSE tasks in realistic open scenarios." @default.
- W4387609344 created "2023-10-14" @default.
- W4387609344 creator A5027067675 @default.
- W4387609344 creator A5035339773 @default.
- W4387609344 creator A5066367494 @default.
- W4387609344 creator A5074337782 @default.
- W4387609344 date "2023-01-01" @default.
- W4387609344 modified "2023-10-14" @default.
- W4387609344 title "Contrastive Learning for Target Speaker Extraction with Attention-based Fusion" @default.
- W4387609344 doi "https://doi.org/10.1109/taslp.2023.3324550" @default.
- W4387609344 hasPublicationYear "2023" @default.
- W4387609344 type Work @default.
- W4387609344 citedByCount "0" @default.
- W4387609344 crossrefType "journal-article" @default.
- W4387609344 hasAuthorship W4387609344A5027067675 @default.
- W4387609344 hasAuthorship W4387609344A5035339773 @default.
- W4387609344 hasAuthorship W4387609344A5066367494 @default.
- W4387609344 hasAuthorship W4387609344A5074337782 @default.
- W4387609344 hasConcept C133892786 @default.
- W4387609344 hasConcept C138885662 @default.
- W4387609344 hasConcept C149838564 @default.
- W4387609344 hasConcept C153180895 @default.
- W4387609344 hasConcept C154945302 @default.
- W4387609344 hasConcept C162324750 @default.
- W4387609344 hasConcept C187736073 @default.
- W4387609344 hasConcept C199360897 @default.
- W4387609344 hasConcept C2776401178 @default.
- W4387609344 hasConcept C2779843651 @default.
- W4387609344 hasConcept C2780451532 @default.
- W4387609344 hasConcept C28490314 @default.
- W4387609344 hasConcept C41008148 @default.
- W4387609344 hasConcept C41608201 @default.
- W4387609344 hasConcept C41895202 @default.
- W4387609344 hasConcept C97931131 @default.
- W4387609344 hasConceptScore W4387609344C133892786 @default.
- W4387609344 hasConceptScore W4387609344C138885662 @default.
- W4387609344 hasConceptScore W4387609344C149838564 @default.
- W4387609344 hasConceptScore W4387609344C153180895 @default.
- W4387609344 hasConceptScore W4387609344C154945302 @default.
- W4387609344 hasConceptScore W4387609344C162324750 @default.
- W4387609344 hasConceptScore W4387609344C187736073 @default.
- W4387609344 hasConceptScore W4387609344C199360897 @default.
- W4387609344 hasConceptScore W4387609344C2776401178 @default.
- W4387609344 hasConceptScore W4387609344C2779843651 @default.
- W4387609344 hasConceptScore W4387609344C2780451532 @default.
- W4387609344 hasConceptScore W4387609344C28490314 @default.
- W4387609344 hasConceptScore W4387609344C41008148 @default.
- W4387609344 hasConceptScore W4387609344C41608201 @default.
- W4387609344 hasConceptScore W4387609344C41895202 @default.
- W4387609344 hasConceptScore W4387609344C97931131 @default.
- W4387609344 hasLocation W43876093441 @default.
- W4387609344 hasOpenAccess W4387609344 @default.
- W4387609344 hasPrimaryLocation W43876093441 @default.
- W4387609344 hasRelatedWork W1493012537 @default.
- W4387609344 hasRelatedWork W1521049138 @default.
- W4387609344 hasRelatedWork W2103897043 @default.
- W4387609344 hasRelatedWork W2125642021 @default.
- W4387609344 hasRelatedWork W2162158162 @default.
- W4387609344 hasRelatedWork W2206035908 @default.
- W4387609344 hasRelatedWork W3148366653 @default.
- W4387609344 hasRelatedWork W4247736853 @default.
- W4387609344 hasRelatedWork W4384929466 @default.
- W4387609344 hasRelatedWork W2175373321 @default.
- W4387609344 isParatext "false" @default.
- W4387609344 isRetracted "false" @default.
- W4387609344 workType "article" @default.