A Modified Isomap Approach to Manifold Learning in Word Spotting.
Sebastian SudholtGernot A. FinkPublished in: GCPR (2015)
Keyphrases
- manifold learning
- word spotting
- nonlinear dimensionality reduction
- dimensionality reduction
- laplacian eigenmaps
- low dimensional
- feature mapping
- dimension reduction
- manifold learning algorithm
- high dimensional
- semi supervised
- subspace learning
- locally linear embedding
- text processing
- nonlinear dimension reduction
- dynamic time warping
- document images
- feature extraction
- optical character recognition
- high dimensional data
- neighborhood graph
- handwritten documents
- geodesic distance
- manifold structure
- handwriting recognition
- low dimensional manifolds
- riemannian manifolds
- unsupervised learning
- feature space
- pattern recognition
- cluster analysis
- data mining