An Algorithm-Hardware Co-Optimized Framework for Accelerating N: M Sparse Transformers.
Chao FangAojun ZhouZhongfeng WangPublished in: CoRR (2022)
Keyphrases
- computational complexity
- experimental evaluation
- np hard
- bayesian framework
- detection algorithm
- learning algorithm
- probabilistic model
- objective function
- computational cost
- preprocessing
- sparse matrix
- convergence rate
- dynamic programming
- cost function
- improved algorithm
- search space
- worst case
- particle swarm optimization
- real time
- times faster
- optimization algorithm
- simulated annealing
- high accuracy
- significant improvement
- input data
- similarity measure
- image processing
- optimal solution
- feature selection
- neural network
- recognition algorithm
- k means
- multiple instance learning
- hardware architecture