Improving RNN Transducer with Target Speaker Extraction and Neural Uncertainty Estimation.
Jiatong ShiChunlei ZhangChao WengShinji WatanabeMeng YuDong YuPublished in: ICASSP (2021)
Keyphrases
- recurrent neural networks
- nearest neighbor
- neural network
- neural model
- information extraction
- audio visual
- estimation accuracy
- decision theory
- speaker recognition
- speaker verification
- accurate estimation
- network architecture
- bio inspired
- neural fuzzy
- pattern recognition
- moving target
- automatic extraction
- mobile robot
- learning rules
- data sets
- automatic speech recognition
- estimation algorithm
- learning algorithm
- genetic algorithm
- feature space
- speech recognition
- back propagation