Multimodal Approach of Speech Emotion Recognition Using Multi-Level Multi-Head Fusion Attention-Based Recurrent Neural Network.
Ngoc-Huynh HoHyung-Jeong YangSoo-Hyung KimGueesang LeePublished in: IEEE Access (2020)
Keyphrases
- recurrent neural networks
- multimodal fusion
- audio visual
- multimodal interfaces
- emotion recognition
- neural network
- multi modal
- human computer interaction
- feed forward
- recurrent networks
- reservoir computing
- complex valued
- hidden layer
- artificial neural networks
- multimodal interaction
- speech recognition
- facial expressions
- facial animation
- emotional state
- information fusion
- text to speech synthesis
- nonlinear dynamic systems
- neural model
- feedforward neural networks
- emotional speech
- autistic children
- long short term memory
- real time
- affect detection
- multimodal biometrics
- multi stream
- hand movements
- text input
- neural network structure
- echo state networks
- text to speech
- viterbi algorithm
- automatic speech recognition
- speech signal