∃stan
Results
2
issues of
∃stan
In the paper, the author say "we first train the network with MAE loss and then **fine-tune** it by MSE loss.",but I cannot find the clue about how to fine-tune...
我看文中好几处提到automatically：“In the inference stage, FADB **automatically** assigns higher frequency regions to heavier branches”，论文里好像也没说是如何自动选择分支的，很疑惑。。。