GLORIA

GEOMAR Library Ocean Research Information Access

X

Your email was sent successfully. Check your inbox.

An error occurred while sending the email. Please try again.

Proceed reservation?

Export
  • 1
    Online Resource
    Online Resource
    Att-TasNet: Attending to Encodings in Time-Domain Audio Speech Separation of Noisy, Reverberant Speech Mixtures
    Frontiers Media SA ; 2022
    In:  Frontiers in Signal Processing Vol. 2 ( 2022-5-11)
    Details
    In: Frontiers in Signal Processing, Frontiers Media SA, Vol. 2 ( 2022-5-11)
    Abstract: Separation of speech mixtures in noisy and reverberant environments remains a challenging task for state-of-the-art speech separation systems. Time-domain audio speech separation networks (TasNets) are among the most commonly used network architectures for this task. TasNet models have demonstrated strong performance on typical speech separation baselines where speech is not contaminated with noise. When additive or convolutive noise is present, performance of speech separation degrades significantly. TasNets are typically constructed of an encoder network, a mask estimation network and a decoder network. The design of these networks puts the majority of the onus for enhancing the signal on the mask estimation network when used without any pre-processing of the input data or post processing of the separation network output data. Use of multihead attention (MHA) is proposed in this work as an additional layer in the encoder and decoder to help the separation network attend to encoded features that are relevant to the target speakers and conversely suppress noisy disturbances in the encoded features. As shown in this work, incorporating MHA mechanisms into the encoder network in particular leads to a consistent performance improvement across numerous quality and intelligibility metrics on a variety of acoustic conditions using the WHAMR corpus, a data-set of noisy reverberant speech mixtures. The use of MHA is also investigated in the decoder network where it is demonstrated that smaller performance improvements are consistently gained within specific model configurations. The best performing MHA models yield a mean 0.6 dB scale invariant signal-to-distortion (SISDR) improvement on noisy reverberant mixtures over a baseline 1D convolution encoder. A mean 1 dB SISDR improvement is observed on clean speech mixtures.
    Type of Medium: Online Resource
    ISSN: 2673-8198
    URL: Article
    DOI: 10.3389/frsip.2022.856968
    Language: Unknown
    Publisher: Frontiers Media SA
    Publication Date: 2022
    detail.hit.zdb_id: 3110502-6
    Location Call Number Limitation Availability
    BibTip Others were also interested in ...
    Online Resource
  • 2
    Online Resource
    Online Resource
    Incidence and predictors of hemodynamic compromise due to high-grade AV block after TAVI
    Frontiers Media SA ; 2023
    In:  Frontiers in Cardiovascular Medicine Vol. 10 ( 2023-6-6)
    Details
    In: Frontiers in Cardiovascular Medicine, Frontiers Media SA, Vol. 10 ( 2023-6-6)
    Abstract: High-grade AV block (HAVB) is the most frequent adverse event after transcatheter aortic valve implantation (TAVI). In rare cases, HAVB is associated with hemodynamic compromise (HC) followed by syncope or application of cardiopulmonary resuscitation (CPR), but data on this severe complication are scarce. The aim of the present study was to investigate the incidence and predictors of HC due to HAVB in patients undergoing TAVI. Methods In this retrospective analysis of 4,602 TAVI cases between 2010 and 2022, 466 developed HAVB. Baseline characteristics and procedural and postprocedural findings were compared for patients with HC versus those without. Univariate and multivariable regression analyses were used to investigate independent predictors of HC. Results Forty-nine of 466 patients (10.5%) had HC due to HAVB after TAVI. Patients with HC had a longer hospital stay [10 (8–13) vs. 13 (9–18) days; p   & lt; 0.001], more frequent peripheral artery disease (PAD) (28.6% vs. 15.1%; p  = 0.016), and lower hemoglobin levels [11.8 (±) vs. 12.5 (±) g/dl; p  = 0.006]. In the HC group, HAVB onset post-TAVI was delayed compared with the non-HC group [2 (1–4) vs. 1 (0–3) days; p   & lt; 0.001]. Before HAVB onset, patients in the HC group more frequently developed post-TAVI delirium [18 (4.6%) vs. 11 (25.0%); p   & lt; 0.001]. In univariate regression analysis, PAD, hemoglobin, procedural time, contrast agent volume, and post-TAVI delirium were significant predictors of HC. After adjustment, only post-TAVI delirium and contrast agent volume remained independent predictors [OR 3.22 (95% CI: 1.05–9.89); p  = 0.042 and OR: 1.01 (95% CI: 1.0–1.01); p  = 0.04, respectively]. Conclusion HC due to HAVB after TAVI occurred in over 10% of cases. Development of post-TAVI delirium and contrast agent volume are independent predictors of this severe complication.
    Type of Medium: Online Resource
    ISSN: 2297-055X
    URL: Article
    DOI: 10.3389/fcvm.2023.1161871
    Language: Unknown
    Publisher: Frontiers Media SA
    Publication Date: 2023
    detail.hit.zdb_id: 2781496-8
    Location Call Number Limitation Availability
    BibTip Others were also interested in ...
    Online Resource
Close ⊗
This website uses cookies and the analysis tool Matomo. More information can be found here...