Multi-scale Residual Network for Image Super-Resolution

By Juncheng Li, Faming Fang, Kangfu Mei, Guixu Zhang. European Conference on Computer Vision, 2018

827 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/7a71941e60894ae7e1f5af8e79c37cec6cd6c6ad

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Cross-level Feature Aggregation Network for Polyp Segmentation (2023, background): “For instance, Li et al. [25] proposed to utilize different sizes of convolution kernels to adaptively detect multi-scale image features for image super-resolution.”
Deep Learning for Image Super-Resolution: A Survey (2019, methodology): “Motivated by inception module [100], the MSRN [98] adopts a new block for multi-scale feature extraction, as Fig.”
Learning Enriched Features for Fast Image Restoration and Enhancement (2022, neutral): “[65], [66], dense connections [7], [67], [68], attention mechanisms [69], [70], [71], multi-branch learning [66], [72], [73], [74], and generative adversarial networks (GANs) [68], [75], [76], [77].”
Learning Enriched Features for Real Image Restoration and Enhancement (2020, background): “Other networks designed to perform SR include recursive learning [57,41,3], progressive reconstruction [105,60], dense connections [99,104,127], attention mechanisms [125,23,126], multi-branch learning [60,66,22,64], and generative adversarial networks (GANs) [104,76,87,63].”
CLEARER: Multi-Scale Neural Architecture Search for Image Restoration (2020, background): “In these handcrafted models, multi-scale architectures [16, 37, 33, 35, 39, 42] have played a significant role in improving the performance, of which the basic idea is to fuse features with different resolutions/scales.”
Camouflaged Object Detection with Feature Decomposition and Edge Reconstruction (2023, methodology): “Following [21, 26], we first apply a residual block for texture preservation, consisting of a 3 × 3 convolution layer, batch normalization (BN) [12], and ReLU.”
Underwater Image Enhancement Using a Multiscale Dense Generative Adversarial Network (2020, neutral): “the underutilization of feature maps [30].”
DARTS-: Robustly Stepping out of Performance Collapse Without Indicators (2020, background): “Multiple works (Ren et al., 2015; Wei et al., 2017; Tai et al., 2017; Li et al., 2018b) also empirically demonstrate that deep residual network can achieve better performance on various tasks.”
SRDiff: Single Image Super-Resolution with Diffusion Probabilistic Models (2021, positive): “[29] proposed a multi-scale residual block (MSRB), which was capable of adaptively detecting and fusing image features at different scales and obtained accurate SR images efﬁ-ciently.”
MADNet: A Fast and Lightweight Network for Single-Image Super Resolution (2020, methodology): “Recently, Li et al. [26] introduced a multiscale residual network to exploit the image features to achieve a signiﬁcant performance gain for image super resolution.”
Transformer for Single Image Super-Resolution (2021, background): “To address this issue, numerous deep neural networks have been proposed [10, 13, 18, 21, 22, 26, 39, 40, 45].”
Deep learning methods for super-resolution reconstruction of turbulent flows (2020, methodology): “The EMSRB is developed from the multi-scale residual block (MSRB), 33 which has been shown as an efficient block to detect image features at different scales.”
Interpreting Super-Resolution Networks with Local Attribution Maps (2020, neutral): “[31], RCAN, RRDBNet, CARN, SAN, IMDN [25], RNAN [60], RFDN [34], PAN [62], DRLN, and CSNLN [37].”
Deep Learning on Image Denoising: An overview (2019, background): “ However, these networks were not easy to add new plug-in units, which limited their applications in the real world [52]. Base on the reasons above, convolutional neural networks (CNNs) were proposed [127, 108]. The CNN as well as LeNet had a real-world application in hand-written digit recognition ”
Efficient Transformer for Single Image Super-Resolution (2021, background): “For example, MSRN [9] proposes a multi-scale block to adaptively detect the image features in different scales.”