Engineering
Magnitude Spectrum
100%
Recurrent
79%
Transfer Learning
69%
Stream Processing
61%
Dimensionality
38%
Phase Spectrum
30%
Fourier Transform
30%
Related Attribute
30%
Imaginary Part
30%
Convolutional Neural Network
30%
Real Part
30%
Base Line
30%
Building Acoustics
30%
Major Problem
30%
Error Rate
30%
Confusion Matrix
15%
Detection Task
10%
Classification Task
10%
Fusion Scheme
10%
Severity Level
10%
Single Stream
7%
Speech Signal
7%
Acoustic Feature
7%
Convolutional Layer
7%
Classification Accuracy
7%
End System
7%
Recognizer
7%
Conformers
7%
Regularization
7%
Metrics
7%
Computer Science
Dysarthric Speech Recognition
92%
Magnitude Spectrum
48%
speech corpus
41%
Information Stream
41%
Speaker Variability
41%
Transfer Learning
41%
Stream Processing
41%
Preprocessing
41%
Mel-Frequency Cepstral Coefficients
41%
Convolutional Layer
38%
Fourier Transform
30%
Imaginary Part
30%
Convolutional Neural Network
30%
Recurrent Layer
30%
Fully Connected Layer
17%
Data Augmentation
14%
Feature Normalization
10%
Dimensionality Reduction
10%
Convolutional Filter
10%
Performance Gain
10%
Related Attribute
10%
Convolutional Network
7%
Information Loss
7%
Postprocessing
7%
Classification Accuracy
7%
