A Two-Stage CNN with Feature Reduction for Speech-Aware Binaural DOA Estimation.
Reza VarzandehSimon DocloVolker HohmannPublished in: EUSIPCO (2023)
Keyphrases
- doa estimation
- feature reduction
- sound source
- speech signal
- audio visual
- direction of arrival
- speech recognition
- rough sets
- text categorization
- random forest
- reduction method
- feature set
- feature selection
- canonical correlation analysis
- linear discriminant analysis
- texture segmentation
- classification accuracy
- knn
- hidden markov models
- multi modal
- support vector machine
- rough set theory
- non stationary
- principal component analysis
- viterbi algorithm
- classification algorithm
- language model
- face recognition
- machine learning
- data sets
- k nearest neighbor
- text classification
- least squares