search

GuoShi28 / CBDNet

Code for "Toward Convolutional Blind Denoising of Real Photographs", CVPR 2019
Apache License 2.0
501 stars 92 forks source link

exe time #15

Open shenyingying opened 5 years ago

shenyingying commented 5 years ago

hi @GuoShi28 your paper said CBDNet takes about 0.4s to process an 512 × 512 image, i was consider in what platform it execute? have you comparison the execute time between CBDnet and BM3D? ths a lot

GuoShi28 commented 5 years ago

hi, I test using matconvnet using 1080Ti.

shenyingying commented 5 years ago

hi, I test using matconvnet using 1080Ti.

@GuoShi28 ths for your replay ,i was wonder how compare the execute time between CBDnet and FFDnet ,that paper give the result i want to transplant denoise algorithm to embedded platform,is there any advice your gave?ths a lot.

GuoShi28 commented 5 years ago

Actually, you can directly compare. CBDNet is much slower for the following reasons:(1) deeper network, UNet structure. (2) ffdnet uses downsampling first to further save running time. Suggestions: I don't have the experience to transplant algorithm to embedded platform. Maybe you can also refer the downsampling operator (e.g., ffdnet) and some traditional speed-up network structures (e.g., mobilenet v2).

shenyingying commented 5 years ago

@GuoShi28 ths a lot!