Feature space mutual information in speech-video sequences.
Torsten ButzJean-Philippe ThiranPublished in: ICME (2) (2002)
Keyphrases
- mutual information
- feature space
- video sequences
- feature selection
- information theoretic
- speech recognition
- image registration
- canonical correlations
- similarity measure
- high dimensional
- training samples
- feature vectors
- principal component analysis
- conditional mutual information
- medical image registration
- speech signal
- frame rate
- video frames
- object tracking
- mean shift
- image sequences
- motion estimation
- audio visual
- video camera
- high dimensionality
- multimodal image registration
- motion vectors
- dimensionality reduction
- information theoretic measures
- classification accuracy
- low dimensional
- kernel function
- camera motion
- training set
- moving objects
- facial motion
- information gain
- input space
- facial expressions
- image representation
- image retrieval
- inter frame
- kernel methods
- data points
- automatic speech recognition
- visual tracking
- video data
- input data
- support vector machine
- feature extraction
- spoken language
- feature set
- text to speech
- mutual information maximization
- speech synthesis
- high dimensional feature space
- dialogue system
- dissimilarity measure
- human motion
- linear discriminant analysis
- computer vision