Hierarchical Attention-Based Contextual Biasing For Personalized Speech Recognition Using Neural Transducers.
Sibo TongPhilip HardingSimon WieslerPublished in: ASRU (2023)
Keyphrases
- neural network
- network architecture
- speech recognition
- contextual information
- speech signal
- e learning
- coarse to fine
- visual attention
- finite state automata
- context sensitive
- text to speech
- finite automata
- context aware
- user modeling
- endpoint detection
- audio stream
- recognition engine
- neural fuzzy
- focus of attention
- action selection
- bio inspired
- adaptive learning
- reinforcement learning
- hidden markov models
- audio visual
- language model
- hierarchical structure