WinoNN: Optimizing FPGA-Based Convolutional Neural Network Accelerators Using Sparse Winograd Algorithm.
Xuan WangChao WangJing CaoLei GongXuehai ZhouPublished in: IEEE Trans. Comput. Aided Des. Integr. Circuits Syst. (2020)
Keyphrases
- dynamic programming
- detection algorithm
- optimization algorithm
- k means
- experimental evaluation
- hardware implementation
- computational complexity
- preprocessing
- theoretical analysis
- matching algorithm
- objective function
- improved algorithm
- significant improvement
- cost function
- linear programming
- high accuracy
- computational cost
- pairwise
- learning algorithm
- convergence rate
- parallel implementation
- tree structure
- detection method
- expectation maximization
- optimal solution
- feature selection
- computer vision