目标检测之损失函数smooth L1 Loss 均方误差MSE与L2 Loss MSE L2 Loss 平均绝对误差MAE与L1 Loss MAE L1 Loss smooth L1 Loss 均方误差MSE与L2 Loss MSE M S E = ∑ i = i n ( y i − y i p ) 2 n MSE=\frac{ {\sum_{i=i}^{n} {(y_i - {y_i}^p)}^2}}{n} MSE=n∑i=in(yi−yip)