Matches in SemOpenAlex for { <https://semopenalex.org/work/W4319862480> ?p ?o ?g. }
- W4319862480 abstract "Recent work on self-supervised pre-training focus on leveraging large-scale unlabeled speech data to build robust end-to-end (E2E) acoustic models (AM) that can be later fine-tuned on downstream tasks e.g., automatic speech recognition (ASR). Yet, few works investigated the impact on performance when the data properties substantially differ between the pre-training and fine-tuning phases, termed domain shift. We target this scenario by analyzing the robustness of Wav2Vec 2.0 and XLS-R models on downstream ASR for a completely unseen domain, air traffic control (ATC) communications. We benchmark these two models on several open-source and challenging ATC databases with signal-to-noise ratio between 5 to 20 dB. Relative word error rate (WER) reductions between 20% to 40% are obtained in comparison to hybrid-based ASR baselines by only fine-tuning E2E acoustic models with a smaller fraction of labeled data. We analyze WERs on the low-resource scenario and gender bias carried by one ATC dataset." @default.
- W4319862480 created "2023-02-11" @default.
- W4319862480 creator A5003975051 @default.
- W4319862480 creator A5021226188 @default.
- W4319862480 creator A5065786740 @default.
- W4319862480 creator A5066348116 @default.
- W4319862480 creator A5076409146 @default.
- W4319862480 creator A5076763342 @default.
- W4319862480 creator A5078502662 @default.
- W4319862480 creator A5084509592 @default.
- W4319862480 creator A5085223859 @default.
- W4319862480 date "2023-01-09" @default.
- W4319862480 modified "2023-10-17" @default.
- W4319862480 title "How Does Pre-Trained Wav2Vec 2.0 Perform on Domain-Shifted Asr? an Extensive Benchmark on Air Traffic Control Communications" @default.
- W4319862480 cites W1540624770 @default.
- W4319862480 cites W2127141656 @default.
- W4319862480 cites W2147590749 @default.
- W4319862480 cites W2514741789 @default.
- W4319862480 cites W2563053575 @default.
- W4319862480 cites W2726599793 @default.
- W4319862480 cites W2903601559 @default.
- W4319862480 cites W2916984808 @default.
- W4319862480 cites W2936774411 @default.
- W4319862480 cites W2973049979 @default.
- W4319862480 cites W2979826702 @default.
- W4319862480 cites W3007073761 @default.
- W4319862480 cites W3015356564 @default.
- W4319862480 cites W3096806943 @default.
- W4319862480 cites W3102342027 @default.
- W4319862480 cites W3109258989 @default.
- W4319862480 cites W3133953372 @default.
- W4319862480 cites W3160641957 @default.
- W4319862480 cites W3163203022 @default.
- W4319862480 cites W3196895794 @default.
- W4319862480 cites W3197876970 @default.
- W4319862480 cites W3198484663 @default.
- W4319862480 cites W3198744226 @default.
- W4319862480 cites W3198771897 @default.
- W4319862480 cites W3206559778 @default.
- W4319862480 cites W4221140371 @default.
- W4319862480 cites W4221147513 @default.
- W4319862480 cites W4225294518 @default.
- W4319862480 doi "https://doi.org/10.1109/slt54892.2023.10022724" @default.
- W4319862480 hasPublicationYear "2023" @default.
- W4319862480 type Work @default.
- W4319862480 citedByCount "5" @default.
- W4319862480 countsByYear W43198624802023 @default.
- W4319862480 crossrefType "proceedings-article" @default.
- W4319862480 hasAuthorship W4319862480A5003975051 @default.
- W4319862480 hasAuthorship W4319862480A5021226188 @default.
- W4319862480 hasAuthorship W4319862480A5065786740 @default.
- W4319862480 hasAuthorship W4319862480A5066348116 @default.
- W4319862480 hasAuthorship W4319862480A5076409146 @default.
- W4319862480 hasAuthorship W4319862480A5076763342 @default.
- W4319862480 hasAuthorship W4319862480A5078502662 @default.
- W4319862480 hasAuthorship W4319862480A5084509592 @default.
- W4319862480 hasAuthorship W4319862480A5085223859 @default.
- W4319862480 hasBestOaLocation W43198624802 @default.
- W4319862480 hasConcept C103824480 @default.
- W4319862480 hasConcept C104317684 @default.
- W4319862480 hasConcept C127413603 @default.
- W4319862480 hasConcept C13280743 @default.
- W4319862480 hasConcept C146978453 @default.
- W4319862480 hasConcept C154945302 @default.
- W4319862480 hasConcept C166961238 @default.
- W4319862480 hasConcept C185592680 @default.
- W4319862480 hasConcept C185798385 @default.
- W4319862480 hasConcept C205649164 @default.
- W4319862480 hasConcept C28490314 @default.
- W4319862480 hasConcept C31972630 @default.
- W4319862480 hasConcept C40969351 @default.
- W4319862480 hasConcept C41008148 @default.
- W4319862480 hasConcept C51632099 @default.
- W4319862480 hasConcept C55493867 @default.
- W4319862480 hasConcept C63479239 @default.
- W4319862480 hasConceptScore W4319862480C103824480 @default.
- W4319862480 hasConceptScore W4319862480C104317684 @default.
- W4319862480 hasConceptScore W4319862480C127413603 @default.
- W4319862480 hasConceptScore W4319862480C13280743 @default.
- W4319862480 hasConceptScore W4319862480C146978453 @default.
- W4319862480 hasConceptScore W4319862480C154945302 @default.
- W4319862480 hasConceptScore W4319862480C166961238 @default.
- W4319862480 hasConceptScore W4319862480C185592680 @default.
- W4319862480 hasConceptScore W4319862480C185798385 @default.
- W4319862480 hasConceptScore W4319862480C205649164 @default.
- W4319862480 hasConceptScore W4319862480C28490314 @default.
- W4319862480 hasConceptScore W4319862480C31972630 @default.
- W4319862480 hasConceptScore W4319862480C40969351 @default.
- W4319862480 hasConceptScore W4319862480C41008148 @default.
- W4319862480 hasConceptScore W4319862480C51632099 @default.
- W4319862480 hasConceptScore W4319862480C55493867 @default.
- W4319862480 hasConceptScore W4319862480C63479239 @default.
- W4319862480 hasLocation W43198624801 @default.
- W4319862480 hasLocation W43198624802 @default.
- W4319862480 hasOpenAccess W4319862480 @default.
- W4319862480 hasPrimaryLocation W43198624801 @default.
- W4319862480 hasRelatedWork W105527591 @default.
- W4319862480 hasRelatedWork W2011731246 @default.
- W4319862480 hasRelatedWork W2152960771 @default.
- W4319862480 hasRelatedWork W2170239774 @default.