2021
DOI: 10.1016/j.knosys.2020.106547
|View full text |Cite
|
Sign up to set email alerts
|

Automated accurate speech emotion recognition system using twine shuffle pattern and iterative neighborhood component analysis techniques

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

2
33
0

Year Published

2021
2021
2024
2024

Publication Types

Select...
4
2
1

Relationship

1
6

Authors

Journals

citations
Cited by 102 publications
(35 citation statements)
references
References 76 publications
2
33
0
Order By: Relevance
“…A classifier is chosen as the error/loss value calculator, the feature vector with the minimum error value is chosen as the optimum feature vector. INCA selects a variant number of features for variant problems [ 41 ]. In this study, we selected the feature range from 50 to 1000 .…”
Section: Methodsmentioning
confidence: 99%
“…A classifier is chosen as the error/loss value calculator, the feature vector with the minimum error value is chosen as the optimum feature vector. INCA selects a variant number of features for variant problems [ 41 ]. In this study, we selected the feature range from 50 to 1000 .…”
Section: Methodsmentioning
confidence: 99%
“…Nowadays, an optimal feature selection and a better machine or classifier are challenging tasks for a robust SER system 32 . For optimal features, the researchers have utilized advanced techniques and deep learning approaches for an emotional speech feature representations due to enormous achievements in different fields for recognition tasks 12 . Hence, the researchers have been inspired by the performances of the deep learning approaches, so they have developed various techniques for the SER and have increased the level of accuracy 33 .…”
Section: Literature Reviewmentioning
confidence: 99%
“…32 For optimal features, the researchers have utilized advanced techniques and deep learning approaches for an emotional speech feature representations due to enormous achievements in different fields for recognition tasks. 12 Hence, the researchers have been inspired by the performances of the deep learning approaches, so they have developed various techniques for the SER and have increased the level of accuracy. 33 Similarly, 34 developed a method to recognize the emotions from a speech spectrogram using the CNN model.…”
Section: Literature Reviewmentioning
confidence: 99%
See 1 more Smart Citation
“…Compared with traditional features [13], log-mel spectrograms try to match human hearing by preserving both the frequency domain as well as the time domain information [14]. It is worth noting that with the introduction of the Attention Mechanism structure [15], which includes various variants such as the Transformer, good results have been achieved in both classification and dimensionality tasks of SER [16].…”
Section: Introductionmentioning
confidence: 99%