Ensemble integration of calibrated speaker localization and statistical speech detection in domestic environments.
Yuuki TachiokaTomohiro NaritaShinji WatanabeJonathan Le RouxPublished in: HSCMA (2014)
Keyphrases
- speech recognition
- audio visual
- automatic speech recognition
- speaker verification
- activity detection
- speaker recognition
- accurate localization
- speaker dependent
- neural network
- detection algorithm
- detection method
- speech signal
- speaker identification
- automatic detection
- speech synthesis
- speaker diarization
- multi modal
- training set
- vocal tract
- prosodic features
- automatic speech recognition systems
- random forests
- object detection
- united states
- false alarms
- anomaly detection
- feature selection
- face recognition
- hidden markov models
- statistical analysis
- dynamic environments
- detection rate
- noisy environments
- acoustic features
- reliable detection
- ensemble methods
- ensemble learning
- broadcast news
- voice activity detection