On the Use of Self-Supervised Pre-Trained Acoustic and Linguistic Features for Continuous Speech Emotion Recognition.
Manon MacaryMarie TahonYannick EstèveAnthony RousseauPublished in: SLT (2021)
Keyphrases
- linguistic features
- pre trained
- speech emotion recognition
- training data
- named entities
- structural features
- text classification
- semantic features
- training examples
- linguistic knowledge
- feature set
- part of speech
- named entity recognition
- control signals
- news stories
- data mining
- image features
- translation model
- probabilistic model
- active learning
- feature space
- information retrieval