Significance of Dimensionality Reduction in CNN-Based Vowel Classification from Imagined Speech Using Electroencephalogram Signals.
Oindrila BanerjeeD. GovindAkhilesh Kumar DubeySuryakanth V. GangashettyPublished in: SPECOM (2022)
Keyphrases
- eeg data
- eeg signals
- dimensionality reduction
- acoustic signals
- fundamental frequency
- feature extraction
- pattern recognition
- speech signal
- spectral features
- high dimensionality
- brain computer interface
- signal processing
- speech recognition
- independent component analysis
- pattern classification
- dimensionality reduction methods
- speech synthesis
- acoustic signal
- feature selection
- feature space
- text classification
- high dimensional
- principal component analysis
- image classification
- blind source separation
- formant frequencies
- feature vectors
- text to speech
- support vector machine svm
- support vector
- single trial
- support vector machine
- machine learning
- convolutional neural network
- pattern recognition and machine learning
- prosodic features
- brain activity
- classification accuracy
- vocal tract
- low dimensional
- supervised dimensionality reduction
- decision trees
- spectral analysis