Matches in SemOpenAlex for { <https://semopenalex.org/work/W4307320512> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4307320512 abstract "Most automatic speech processing systems register degraded performance when applied to noisy or reverberant speech. But how can one tell whether speech is noisy or reverberant? We propose Brouhaha, a neural network jointly trained to extract speech/non-speech segments, speech-to-noise ratios, and C50room acoustics from single-channel recordings. Brouhaha is trained using a data-driven approach in which noisy and reverberant audio segments are synthesized. We first evaluate its performance and demonstrate that the proposed multi-task regime is beneficial. We then present two scenarios illustrating how Brouhaha can be used on naturally noisy and reverberant data: 1) to investigate the errors made by a speaker diarization model (pyannote.audio); and 2) to assess the reliability of an automatic speech recognition model (Whisper from OpenAI). Both our pipeline and a pretrained model are open source and shared with the speech community." @default.
- W4307320512 created "2022-10-31" @default.
- W4307320512 creator A5005882992 @default.
- W4307320512 creator A5007620149 @default.
- W4307320512 creator A5020182193 @default.
- W4307320512 creator A5039894584 @default.
- W4307320512 creator A5042831132 @default.
- W4307320512 creator A5049821241 @default.
- W4307320512 creator A5053248817 @default.
- W4307320512 creator A5057205161 @default.
- W4307320512 creator A5060751140 @default.
- W4307320512 creator A5079257674 @default.
- W4307320512 date "2022-10-24" @default.
- W4307320512 modified "2023-10-14" @default.
- W4307320512 title "Brouhaha: multi-task training for voice activity detection, speech-to-noise ratio, and C50 room acoustics estimation" @default.
- W4307320512 doi "https://doi.org/10.48550/arxiv.2210.13248" @default.
- W4307320512 hasPublicationYear "2022" @default.
- W4307320512 type Work @default.
- W4307320512 citedByCount "0" @default.
- W4307320512 crossrefType "posted-content" @default.
- W4307320512 hasAuthorship W4307320512A5005882992 @default.
- W4307320512 hasAuthorship W4307320512A5007620149 @default.
- W4307320512 hasAuthorship W4307320512A5020182193 @default.
- W4307320512 hasAuthorship W4307320512A5039894584 @default.
- W4307320512 hasAuthorship W4307320512A5042831132 @default.
- W4307320512 hasAuthorship W4307320512A5049821241 @default.
- W4307320512 hasAuthorship W4307320512A5053248817 @default.
- W4307320512 hasAuthorship W4307320512A5057205161 @default.
- W4307320512 hasAuthorship W4307320512A5060751140 @default.
- W4307320512 hasAuthorship W4307320512A5079257674 @default.
- W4307320512 hasBestOaLocation W43073205121 @default.
- W4307320512 hasConcept C115961682 @default.
- W4307320512 hasConcept C121332964 @default.
- W4307320512 hasConcept C127413603 @default.
- W4307320512 hasConcept C154945302 @default.
- W4307320512 hasConcept C155635449 @default.
- W4307320512 hasConcept C199360897 @default.
- W4307320512 hasConcept C201995342 @default.
- W4307320512 hasConcept C204201278 @default.
- W4307320512 hasConcept C24890656 @default.
- W4307320512 hasConcept C2780451532 @default.
- W4307320512 hasConcept C28490314 @default.
- W4307320512 hasConcept C41008148 @default.
- W4307320512 hasConcept C43521106 @default.
- W4307320512 hasConcept C61328038 @default.
- W4307320512 hasConcept C95851461 @default.
- W4307320512 hasConcept C99498987 @default.
- W4307320512 hasConceptScore W4307320512C115961682 @default.
- W4307320512 hasConceptScore W4307320512C121332964 @default.
- W4307320512 hasConceptScore W4307320512C127413603 @default.
- W4307320512 hasConceptScore W4307320512C154945302 @default.
- W4307320512 hasConceptScore W4307320512C155635449 @default.
- W4307320512 hasConceptScore W4307320512C199360897 @default.
- W4307320512 hasConceptScore W4307320512C201995342 @default.
- W4307320512 hasConceptScore W4307320512C204201278 @default.
- W4307320512 hasConceptScore W4307320512C24890656 @default.
- W4307320512 hasConceptScore W4307320512C2780451532 @default.
- W4307320512 hasConceptScore W4307320512C28490314 @default.
- W4307320512 hasConceptScore W4307320512C41008148 @default.
- W4307320512 hasConceptScore W4307320512C43521106 @default.
- W4307320512 hasConceptScore W4307320512C61328038 @default.
- W4307320512 hasConceptScore W4307320512C95851461 @default.
- W4307320512 hasConceptScore W4307320512C99498987 @default.
- W4307320512 hasLocation W43073205121 @default.
- W4307320512 hasOpenAccess W4307320512 @default.
- W4307320512 hasPrimaryLocation W43073205121 @default.
- W4307320512 hasRelatedWork W1535018509 @default.
- W4307320512 hasRelatedWork W2154353037 @default.
- W4307320512 hasRelatedWork W2418631473 @default.
- W4307320512 hasRelatedWork W2485008119 @default.
- W4307320512 hasRelatedWork W2755891984 @default.
- W4307320512 hasRelatedWork W2760287881 @default.
- W4307320512 hasRelatedWork W2784225896 @default.
- W4307320512 hasRelatedWork W2997245634 @default.
- W4307320512 hasRelatedWork W4307320512 @default.
- W4307320512 hasRelatedWork W642007152 @default.
- W4307320512 isParatext "false" @default.
- W4307320512 isRetracted "false" @default.
- W4307320512 workType "article" @default.