Feature selection based on Multiple Kernel Learning for single-channel sound source localization using the acoustic transfer function.
Ryoichi TakashimaTetsuya TakiguchiYasuo ArikiPublished in: ICASSP (2011)
Keyphrases
- sound source
- multiple kernel learning
- source localization
- single channel
- transfer function
- feature selection
- multi task
- speech signal
- linear combination
- audio visual
- support vector
- dimensionality reduction
- image enhancement
- support vector machine
- model selection
- machine learning
- real environment
- focus of attention
- neural network
- edge detection
- contrast enhancement
- augmented reality
- training data
- dynamic environments
- feature extraction
- multimedia
- multi modal