2022
DOI: 10.1109/access.2022.3176441
|View full text |Cite
|
Sign up to set email alerts
|

Frequency-Based Enhancement Network for Efficient Super-Resolution

Abstract: Recently, deep convolutional neural networks (CNNs) have provided outstanding performance in single image super-resolution (SISR). Despite their remarkable performance, the lack of highfrequency information in the recovered images remains a core problem. Moreover, as the networks increase in depth and width, deep CNN-based SR methods are faced with the challenge of computational complexity in practice. A promising and under-explored solution is to adapt the amount of compute based on the different frequency ba… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

0
3
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
5
1

Relationship

1
5

Authors

Journals

citations
Cited by 7 publications
(3 citation statements)
references
References 70 publications
0
3
0
Order By: Relevance
“…In this section, we compare our FSANet with state-of-the-art methods, including SRCNN [3], FSRCNN [9], VDSR [6], HDRN [48], CARN [49], MemNet [7], IMDN [50], LAPAR-A [51], SRMD [52], A2F-L [53], and FENet [54].…”
Section: Comparison and Visualization Of Psnr And Ssim Resultsmentioning
confidence: 99%
“…In this section, we compare our FSANet with state-of-the-art methods, including SRCNN [3], FSRCNN [9], VDSR [6], HDRN [48], CARN [49], MemNet [7], IMDN [50], LAPAR-A [51], SRMD [52], A2F-L [53], and FENet [54].…”
Section: Comparison and Visualization Of Psnr And Ssim Resultsmentioning
confidence: 99%
“…Also, the feedback mechanism is exploited by Li, Zhen, et al [26] to refine the output of the network. Another mechanism is proposed in FENet by Behjati et al [27] using a frequency-based enhancement network. Luo et al [28] proposed a novel adversarial neural degradation (AND) model for blind image SR to generate a wide range of complex degradation effects that are highly non-linear.…”
Section: A Image Super-resolutionmentioning
confidence: 99%
“…Previous work introduced attention mechanisms [16] [17] into super resolution reconstruction to simulate spatial positions, channels, or the interdependence between the two to solve this problem. Recent frequencybased methods [18] use more complex convolution operations to preserve high frequency information in the network. Although these methods achieve good reconstruction results to a certain extent, the attention-based method does not distinguish between low frequency and high frequency features, while the frequency-based method simply uses the residual branch as high frequency information.…”
Section: Introductionmentioning
confidence: 99%