Training speech emotion classifier without categorical annotations.
Meysam ShamsiMarie TahonPublished in: CoRR (2022)
Keyphrases
- training set
- training process
- emotion recognition
- ground truth labels
- training samples
- training examples
- labeled images
- hearing impaired
- classifier training
- text to speech synthesis
- training data
- supervised training
- decision trees
- feature selection and classifier
- emotional speech
- discriminative classifiers
- text classifiers
- emotional state
- speech recognition
- feature selection
- learning algorithm
- image classification
- active learning
- support vector
- boosted classifiers
- semantic annotation
- classification performances
- metadata
- facial expressions
- test set
- labeled training data
- feature space
- semantic labels
- speaker identification
- training dataset
- categorical data
- data sets
- human computer interaction
- image annotation
- sufficient training data
- audio visual
- lexical features
- broadcast news
- speech signal
- feature set
- multi class
- feature extraction
- neural network