In our work, a combination of MobileNet v2 and depth-wise separable convolution is proposed to reduce the number of model parameters and the ...
確定! 回上一頁