SP-IMC: A Sparsity Aware In-Memory-Computing Macro in 28nm CMOS with Configurable Sparse Representation for Highly Sparse DNN Workloads.
Amitesh SridharanFan ZhangJae-Sun SeoDeliang FanPublished in: CICC (2024)
Keyphrases
- sparse representation
- dictionary learning
- sparse coding
- sparsity constraints
- sparse approximations
- image classification
- compressed sensing
- compressive sensing
- regularized least squares
- high dimensional data
- random projections
- negative matrix factorization
- joint optimization
- signal processing
- sparse reconstruction
- sparse coefficients
- face recognition
- image patches
- image representation
- dimensionality reduction
- nm technology
- cmos technology
- orthogonal matching pursuit
- power consumption
- data points
- pattern recognition
- sparse codes
- silicon on insulator
- low power
- computer vision
- feature extraction
- sparse learning
- training set
- norm minimization
- power dissipation
- basis vectors