search

xiongzhu666 / Sky-Segmentation-and-Post-processing

MIT License
110 stars 23 forks source link

Image resolution #6

Open lknnnli opened 2 years ago

lknnnli commented 2 years ago

Hi Xiong,

Recently I am reading that google paper, and I have a issue want to confirm with you. image As I see from the pipeline, in order to get coarse matte, we need to first down size the input raw image, And I think this down-sized image should be the input of matte inference network( for your case is U2net). and the following upsampled weighted guided filter is used to upsample the low resolution output from u2net to a high resolution same as input.

Am I correct? With warm regards Kangning Li

xiongzhu666 commented 2 years ago

Hi Xiong,

Recently I am reading that google paper, and I have a issue want to confirm with you. image As I see from the pipeline, in order to get coarse matte, we need to first down size the input raw image, And I think this down-sized image should be the input of matte inference network( for your case is U2net). and the following upsampled weighted guided filter is used to upsample the low resolution output from u2net to a high resolution same as input.

Am I correct? With warm regards Kangning Li

Yes

  1. the inference network give a low resolution mask and I resize the mask to original resolution by linear interpolation in opencv.
  2. the weighted guided filter is for refine mask and keep more details. its inputs are orignal RGB image and linear interpolation mask.