Jump Diffusion over Feature Space for Object Recognition.
Sébastien GadatPublished in: SIAM J. Control. Optim. (2008)
Keyphrases
- feature space
- object recognition
- image representation
- computer vision
- feature vectors
- high dimensional
- scene classification
- template matching
- natural images
- mean shift
- object categorization
- dimensionality reduction
- image matching
- classification accuracy
- feature selection
- training samples
- image features
- input space
- object detection
- kernel function
- constrained search
- visual words
- visual recognition
- image retrieval
- high dimensional feature space
- feature extraction
- model construction
- shape recognition
- canonical correlations
- high dimensionality
- object categories
- image understanding
- kernel methods
- d objects
- input data
- principal component analysis
- dimension reduction
- linear discriminant analysis
- graph matching
- dissimilarity measure
- object classification
- biologically motivated
- low dimensional
- support vector machine
- data points
- visual object recognition
- dot product
- riemannian manifolds
- low level
- keypoints
- neural network
- feature matching
- training set