Graph Convolutional Networks based on manifold learning for semi-supervised image classification.
Lucas Pascotti ValemDaniel Carlos Guimarães PedronetteLongin Jan LateckiPublished in: Comput. Vis. Image Underst. (2023)
Keyphrases
- manifold learning
- semi supervised
- image classification
- neighborhood graph
- label propagation
- feature extraction
- graph construction
- subspace learning
- sparse coding
- semi supervised learning
- sparse representation
- diffusion maps
- nonlinear dimensionality reduction
- unlabeled data
- low dimensional
- labeled data
- dimensionality reduction
- pairwise
- image representation
- graph embedding
- dimension reduction
- random walk
- manifold structure
- active learning
- unsupervised learning
- high dimensional data
- supervised learning
- laplacian eigenmaps
- high dimensional
- bag of words
- generative model
- image features
- graph laplacian
- pairwise constraints
- weighted graph
- locally linear embedding
- least squares
- neural network
- geodesic distance
- metric learning
- low rank
- training set