Vision Transformer Compression with Structured Pruning and Low Rank Approximation.
Ankur KumarPublished in: CoRR (2022)
Keyphrases
- low rank approximation
- singular value decomposition
- low rank matrix approximation
- computer vision
- subspace learning
- image compression
- low rank
- compression ratio
- spectral clustering
- latent semantic indexing
- kernel matrix
- compression algorithm
- data dependent
- iterative algorithms
- fault diagnosis
- reconstruction error
- nonnegative matrix factorization
- data sets
- feature vectors
- matrix completion
- feature selection
- image processing
- image segmentation
- linear combination