Abstract: The development of flexible Convolutional Neural Network (CNN) accelerators is critical for large-scale inference and training. Accelerators based on the General Matrix Multiplication (GEMM) ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results