Automatically augmenting an emotion dataset improves classification using audio.
Egor LakomkinCornelius WeberStefan WermterPublished in: EACL (2) (2017)
Keyphrases
- pattern recognition
- emotion recognition
- classification accuracy
- classification systems
- benchmark datasets
- multimedia
- classification scheme
- classification method
- classification algorithm
- uci datasets
- training dataset
- pattern classification
- feature space
- image classification
- support vector machine svm
- unsupervised learning
- machine learning methods
- pre classified
- data sets
- decision trees
- training set
- support vector
- emotion classification
- automatically classify
- neural network
- classification process
- feature extraction
- video sequences
- database
- preprocessing
- support vector machine
- multi class
- decision rules
- class labels
- feature set
- text classification