search

zhengchen1999 / DAT

PyTorch code for our ICCV 2023 paper "Dual Aggregation Transformer for Image Super-Resolution"
Apache License 2.0
386 stars 37 forks source link

Personal Confusion About the paper #5

Open MeycL opened 1 year ago

MeycL commented 1 year ago

Thanks for this wonderful work, I've been reading your article recently, may I ask what is the size of this feature map? Underneath it says . ...Nw x Nw, but I am still vague about the process, can you please help me with this query? image

zhengchen1999 commented 1 year ago

Hi. Thanks for your interest in our work.

The size of the feature map in red box is $\frac{HW}{N_w} \times N_w \times N_w$. For simplicity, we omit $\frac{HW}{N_w}$ with ellipsis, which is the number of windows. More specifically, ${N_w}$ is the window size. For example, if the window is 4 $\times$ 16, ${N_w}=4\times16=64$.