Occlusion Sensitivity Analysis with Augmentation Subspace Perturbation in Deep Feature Space.
Pedro H. V. ValoisKoichiro NiinumaKazuhiro FukuiPublished in: WACV (2024)
Keyphrases
- sensitivity analysis
- feature space
- low dimensional
- principal component analysis
- managerial insights
- kernel based nonlinear
- dimensionality reduction
- feature vectors
- high dimensional
- lower dimensional
- influence diagrams
- variational inequalities
- high dimensionality
- input data
- training samples
- input space
- dot product
- mean shift
- classification accuracy
- feature extraction
- kernel methods
- feature selection
- kernel function
- canonical correlations
- high dimensional data
- training set
- image retrieval
- support vector machine
- linear discriminant analysis
- dimension reduction
- data points
- viewpoint
- subspace methods
- decision variables
- independent components
- kernel trick
- knapsack problem
- linear programming