Neural network vowel-recognition jointly using voice features and mouth shape image.
Jian-Tong WuShinichi TamuraHiroshi MitsumotoHideo KawaiKenji KurosuKozo OkazakiPublished in: Pattern Recognit. (1991)
Keyphrases
- image features
- neural network
- test images
- invariant features
- geometrical features
- low level
- visual appearance
- facial feature points
- multiscale
- image statistics
- image classification
- extracted features
- extracting features
- invariant properties
- geometric shapes
- spatial layout
- pixel intensities
- input image
- target object
- image representation
- image content
- single image
- spatial information
- object shapes
- keypoints
- edge detection
- pattern recognition
- appearance features
- feature vectors
- fourier descriptors
- spatial arrangement
- image segmentation
- image regions
- feature extraction
- high resolution
- moment invariants
- segmentation algorithm
- image retrieval
- facial region
- high dimensional
- shape properties
- computer vision
- segmentation method
- data sets
- video sequences
- similarity measure
- nearest neighbor
- image structure
- shape analysis
- feature selection
- pictorial structure
- pattern classification
- feature set
- intensity histogram
- vowel recognition