TY - GEN
T1 - Efficient Pyramid-Structured Networks Using Channel Separating Gate
AU - Yeo, Yoon Jae
AU - Sagong, Min Cheol
AU - Ko, Sung Jea
AU - Lim, Dong Pan
N1 - Funding Information:
This work was supported by Samsung Electronics Co., Ltd (IO201210-08026-01)
Publisher Copyright:
© 2021 IEEE.
PY - 2021
Y1 - 2021
N2 - Convolutional neural networks (CNNs) have been brought noticeable improvement on numerous image processing applications. In particular, spatial pyramid structures have been widely used, but the number of required parameters increases exponentially as the network deepens. Therefore, we propose a channel separating gate (CSG) that can determine the number of channels in convolutional layers according to the spatial level. For a given spatial level, the CSG separates the channels of feature maps for use at the smaller level. Consequently, the standard non- pooling model can be altered to an efficient pyramid structure by performing the proposed method repeatedly. For single image super-resolution (SR) and deblurring, we demonstrate that our method improves the baseline network in terms of peak signal- to-noise ratio (PSNR) and floating-point operations (FLOPs).
AB - Convolutional neural networks (CNNs) have been brought noticeable improvement on numerous image processing applications. In particular, spatial pyramid structures have been widely used, but the number of required parameters increases exponentially as the network deepens. Therefore, we propose a channel separating gate (CSG) that can determine the number of channels in convolutional layers according to the spatial level. For a given spatial level, the CSG separates the channels of feature maps for use at the smaller level. Consequently, the standard non- pooling model can be altered to an efficient pyramid structure by performing the proposed method repeatedly. For single image super-resolution (SR) and deblurring, we demonstrate that our method improves the baseline network in terms of peak signal- to-noise ratio (PSNR) and floating-point operations (FLOPs).
KW - convolutional neural network (CNN)
KW - deblurring
KW - pyramid structure
KW - super-resolution (SR)
UR - http://www.scopus.com/inward/record.url?scp=85123476951&partnerID=8YFLogxK
U2 - 10.1109/GCCE53005.2021.9621933
DO - 10.1109/GCCE53005.2021.9621933
M3 - Conference contribution
AN - SCOPUS:85123476951
T3 - 2021 IEEE 10th Global Conference on Consumer Electronics, GCCE 2021
SP - 986
EP - 987
BT - 2021 IEEE 10th Global Conference on Consumer Electronics, GCCE 2021
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 10th IEEE Global Conference on Consumer Electronics, GCCE 2021
Y2 - 12 October 2021 through 15 October 2021
ER -