Closed FlyFish-space closed 2 years ago
The nest optimization is not based on a gradient-optimized way. In practice, we manually compute the updated values in each iteration. As a result, we published the optimal values in the static_kernel_size
.
ok,thanks!
hello,I have a question like this ,too. the thresholds r2-4 updated values not use the hyperparameter search method ,while manually Grid search?
Please refer to the Sec.III.C in paper. For convenience, we provide the deblurring results of DPDBlur test dataset (https://drive.google.com/file/d/1xkRaaJbqH6R0Mv42Rea3nB4hkRmgodbj/view?usp=sharing).
ok,thank you!
In Section III C of your paper, the thresholds r2-4 are updated by adaptively learning from a small amountof meta-data. However, I don't find this part in your code. I only find you give four
static_kernel_size
. ”Combined with the model training process, generatingthe defocus mask is a nested optimization“ Where is the nested optimization?