Multi-Stream Convolution-Recurrent Neural Networks Based on Attention Mechanism Fusion for Speech Emotion Recognition.
Huawei TaoLei GengShuai ShanJingchao MaiHongliang FuPublished in: Entropy (2022)
Keyphrases
- recurrent neural networks
- multi stream
- speech emotion recognition
- attention mechanism
- audio visual speech recognition
- visual attention
- audio visual
- hidden markov models
- feed forward
- neural network
- saliency map
- recurrent networks
- echo state networks
- visual attention model
- artificial neural networks
- image processing
- information fusion
- high level
- multiscale