A novel approach for matched reverberant training of HMMs using data pairs.
Armin SehrChristian HofmannRoland MaasWalter KellermannPublished in: INTERSPEECH (2010)
Keyphrases
- data sets
- data analysis
- database
- raw data
- data collection
- image data
- data sources
- data processing
- synthetic data
- training samples
- statistical analysis
- high quality
- knowledge discovery
- prior knowledge
- data distribution
- data quality
- data points
- hidden markov models
- data structure
- labeled data
- training examples
- high dimensional data
- machine learning
- missing data
- experimental data
- statistical methods
- neural network