Inception residual attention network for remote sensing image super-resolution

杨利霞

Inception residual attention network for remote sensingimage super-resolution

2 ^9 k  ]$ D' x& Q  b* _1 `: A- dHow to enhance the spatial resolution for a remote sensing image is
3 T3 m# z2 h2 d) M+ g% U: Zan important issue that we face. Many image super-resolution (SR)
techniques have been proposed for this purpose and deep con
* [0 b' x: R7 U2 r% [* {0 a, Wvolutional neural network (CNN) is the most effective approach in
( }6 q% @, m. {1 b/ C- M# G$ Qrecent years. However, we observe that most CNN-based SR meth
, P; M8 |5 h7 A1 Z" f- ?ods treat low-frequency areas and high-frequency areas equally,
hence hindering the recovery of high-frequency information. In this
0 w$ d5 a) u, V: T3 L/ r6 Jpaper, we propose a network named inception residual attention
network (IRAN) to address this problem. Specifically, we propose
a spatial attention module to make the network adaptively learn
the importance of different spatial areas, so as to pay more atten
9 n" D6 T* f9 |( F# o! ^tion to the areas with high-frequency information. Furthermore, we
; s- C9 Y- O# b8 spresent an inception module to fuse local multilevel features, so as
4 w  y! ~, O. F5 ?to provide richer information for reconstructing detailed textures. In
order to evaluate the effectiveness of the proposed method, a large
number of experiments are performed on UCMerced-LandUse data
set and the results show that the proposed method is superior to
the current state-of-the-art methods in both visual effects and
objective indicators.

* x% I4 I; H8 x  T3 y

" U# E, |. G! Q4 N+ M
7 A7 ^4 s1 F6 M/ Q: E' L. a  r