Embedding stacked bottleneck vocal features in a LSTM architecture for automatic pain level classification during emergency triage.
Fu-Sheng TsaiYi-Ming WengChip-Jin NgChi-Chun LeePublished in: ACII (2017)
Keyphrases
- feature vectors
- classification accuracy
- feature extraction
- classification method
- feature set
- classification process
- feature space
- feature representation
- extracted features
- features extraction
- low level
- classification models
- image features
- support vector machine
- management system
- extracting features
- pattern recognition
- discriminative features
- feature analysis
- statistical classification
- image classification
- text classification
- benchmark datasets
- gender classification
- svm classifier
- feature reduction
- feature subset
- automatic classification
- feature weights
- feature level
- class labels
- feature selection algorithms
- high dimensionality
- support vector machine svm
- co occurrence
- decision making
- feature selection
- machine learning
- classification algorithm
- false positives
- feature values
- neural network
- training set
- support vector
- learning algorithm