1.5 million subspaces of a local feature space for 3D object recognition.
Koichi KiseTakahiro KashiwagiPublished in: ACPR (2011)
Keyphrases
- feature space
- high dimensional
- low dimensional
- canonical correlations
- principal component analysis
- high dimensional feature space
- lower dimensional
- feature vectors
- high dimensional feature spaces
- object recognition
- data points
- d objects
- classification accuracy
- kernel function
- feature extraction
- original data
- basis vectors
- mean shift
- input space
- dimensionality reduction
- high dimensionality
- tens of thousands
- dimension reduction
- training samples
- feature selection
- input data
- real world
- range images
- kernel methods
- image representation
- support vector machine
- data sets
- high dimensional data
- training set
- high quality
- hyperplane
- sparse coding
- viewpoint
- dissimilarity measure
- object class
- feature set
- subspace learning
- linear subspace
- canonical correlation analysis
- similarity measure
- clustering algorithm
- linear discriminant analysis
- nearest neighbor