A subspace-accelerated split Bregman method for sparse data recovery with joint l1-type regularizers.
Valentina De SimoneDaniela di SerafinoMarco ViolaPublished in: CoRR (2019)
Keyphrases
- sparse data
- high dimensional
- split bregman method
- low dimensional
- total variation
- dimension reduction
- dimensionality reduction
- data sets
- high dimensional data
- feature space
- multiresolution
- data points
- higher order
- image classification
- support vector
- clustering method
- image restoration
- deformable models
- data structure
- image sequences
- image processing
- computer vision