2021
DOI: 10.1016/j.neucom.2021.05.065
|View full text |Cite
|
Sign up to set email alerts
|

On combining acoustic and modulation spectrograms in an attention LSTM-based system for speech intelligibility level classification

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
14
0

Year Published

2021
2021
2024
2024

Publication Types

Select...
6

Relationship

1
5

Authors

Journals

citations
Cited by 12 publications
(15 citation statements)
references
References 37 publications
1
14
0
Order By: Relevance
“…In [8], we proposed a SIC system based on per-frame log-mel spectrograms and attention LSTM networks that significantly outperformed a SVM-based system with hand-crafted features. Furthermore, in [11] we showed that this architecture is also suitable for the modeling of per-frame modulation spectrograms [27] and that the combination of logmel and modulation spectrograms into an attentional LSTM framework outperforms the corresponding individual systems.…”
Section: Introductionmentioning
confidence: 90%
See 4 more Smart Citations
“…In [8], we proposed a SIC system based on per-frame log-mel spectrograms and attention LSTM networks that significantly outperformed a SVM-based system with hand-crafted features. Furthermore, in [11] we showed that this architecture is also suitable for the modeling of per-frame modulation spectrograms [27] and that the combination of logmel and modulation spectrograms into an attentional LSTM framework outperforms the corresponding individual systems.…”
Section: Introductionmentioning
confidence: 90%
“…Following our previous works on speech intelligibility classification [8,11], the SIC system is based on long short-term memory networks [42,43].…”
Section: Many-to-one Lstm Networkmentioning
confidence: 99%
See 3 more Smart Citations