TY - GEN
T1 - FBRNN
T2 - 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2020
AU - Lee, Junyeop
AU - Park, Jaihyun
AU - Lee, Kanghyu
AU - Min, Jeongki
AU - Kim, Gwantae
AU - Lee, Bokyeung
AU - Ku, Bonhwa
AU - Han, David K.
AU - Ko, Hanseok
N1 - Funding Information:
This material is based upon work supported by the Air Force Office of Scientific Research under award number FA2386-19-1-4001.
Publisher Copyright:
© 2020 IEEE.
PY - 2020/6
Y1 - 2020/6
N2 - Single image extreme Super Resolution (SR) is a difficult task as scale factor in the order of 10X or greater is typically attempted. For instance, in the case of 16x upscale of an image, a single pixel from a low resolution image gets expanded to a 16x16 image patch. Such attempts often result fuzzy quality and loss in details in reconstructed images. To handle these difficulties, we propose a network architecture composed of a series of connected blocks in recurrent and feedback fashions for enhanced SR reconstruction. By use of recurrent network, an SR image is refined over a sequence of enhancement stages in coarse to fine manner. Additionally, each stage involves back projection of SR image to LR images for continuously being refined during the sequence. According to the preliminary results of NTIRE 2020 Perceptual Extreme SR challenge, our team (KU-ISPLB) secured 6th place by PSNR and 7th place by SSIM among all participants.
AB - Single image extreme Super Resolution (SR) is a difficult task as scale factor in the order of 10X or greater is typically attempted. For instance, in the case of 16x upscale of an image, a single pixel from a low resolution image gets expanded to a 16x16 image patch. Such attempts often result fuzzy quality and loss in details in reconstructed images. To handle these difficulties, we propose a network architecture composed of a series of connected blocks in recurrent and feedback fashions for enhanced SR reconstruction. By use of recurrent network, an SR image is refined over a sequence of enhancement stages in coarse to fine manner. Additionally, each stage involves back projection of SR image to LR images for continuously being refined during the sequence. According to the preliminary results of NTIRE 2020 Perceptual Extreme SR challenge, our team (KU-ISPLB) secured 6th place by PSNR and 7th place by SSIM among all participants.
UR - http://www.scopus.com/inward/record.url?scp=85090155281&partnerID=8YFLogxK
U2 - 10.1109/CVPRW50498.2020.00252
DO - 10.1109/CVPRW50498.2020.00252
M3 - Conference contribution
AN - SCOPUS:85090155281
T3 - IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops
SP - 2021
EP - 2028
BT - Proceedings - 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2020
PB - IEEE Computer Society
Y2 - 14 June 2020 through 19 June 2020
ER -