Adaptive Densely Connected Single Image Super-Resolution

Adaptive Densely Connected Single Image Super-Resolution


For a better performance in single image super-resolution(SISR), we present an image super-resolution algorithm based on adaptive dense connection (ADCSR). The algorithm is divided into two parts: BODY and SKIP. BODY improves the utilization of convolution features through adaptive dense connections. Also, we develop an adaptive sub-pixel reconstruction layer (AFSL) to reconstruct the features of the BODY output. We pre-trained SKIP to make BODY focus on high-frequency feature learning. The comparison of PSNR, SSIM, and visual effects verify the superiority of our method to the state-of-the-art algorithms.


1 Introduction

Single image super-resolution aims at reconstructing an accurate high-resolution image from the low-resolution image. Since deep learning made big progress in the computer version, many SISR algorithms based on deep Convolution Neural Networks (CNN) have been proposed in recent years. The powerful feature representation and end-to-end training skill of CNN makes a huge breakthrough in SISR.

Dong \etal [5] first proposed SRCNN by introducing a three-layer CNN for image SR. Kim et al. increased the number of layers to 20 in VDSR [10] and DRCN [11],making notable improvements over SRCNN. As we all know, the deeper the network is, the more powerful the representation it has. However, with the depth of network grow, gradient disappear and gradient explosion will be the main problem to hinder the performance of the network. This problem was solved when He \etal [6] proposed residual net (ResNet), and Huang \etal [7] proposed dense net (DesNet). Many large scale networks were introduced in SISR, such as SRResNet [14], EDSR [16], SRDenseNet [24], RDN [32] etc.These methods aim at building a deeper network to increase the performance. Other methods such as RCAN [31] and SAN [4] try to learn the correlation of the features in the middle layers.

WDSR [28] allows for better network performance with less computational effort. AWSRN [25] applies an adaptive weighted network. Weight adaptation is achieved by multiplying the residual convolution and the residual hopping by coefficients respectively, and the coefficients can be trained. Since the performance of dense connections is better than the residual [16] [32], we develop an adaptive densely connection method to enhance the efficiency of feature learning. There is a similar global SKIP, a single sub-pixel convolution, in WDSR [28]and AWSRN [25]. Although the SKIP is set to recover low-order frequencies, there is no practical measure to limit its training. We present an adaptive densely connected super-resolution reconstruction algorithm (ADCSR). The algorithm is divided into two parts: BODY and SKIP. BODY is focused on high-frequency information reconstruction through pre-training the SKIP. ADCSR obtained the optimal SISR performance based on bicubic interpolation. There are three main tasks:

(1)WDSR [4]is optimized using adaptive dense connections. Experiments were carried out by initializing the adaptive parameters and optimizing the models. Based on the above efforts, the performance of the network has been greatly improved;

(2)We propose the AFSL model to perform image SR through adaptive sub-pixel convolution;

(3)We develop a method which pre-train SKIP first and then train the entire network at the same time. Thus, the BODY is focused on the reconstruction of high-frequency details to improve network performance.

2 Related Works

SISR has important applications in many fields, such as security and surveillance imaging [33], medical imaging [21], and image generation [9]. The simplest method among them is the interpolation, such as linear interpolation, bicubic interpolation, and so on. This method takes the average of the pixel points in the known LR image as the missing pixel of the HR image. Interpolation works well in the smooth part of the image, but it works poorly in the edge regions, causing ringing and blurring. Additionally, learning-based and reconstruction-based methods are more complex such as sparse coding [27], neighborhood embedded regression [3] [23], random forest [19], etc.

Dong et al. first proposed a Convolutional Neural Network (CNN)-based super-resolution reconstruction network (SRCNN) [5], which performance is better than the most advanced algorithm at the time. Later, Shi et al. proposed a sub-pixel convolution super-resolution reconstruction network [20]. The network contains several convolutional layers to learn LR image features. Reconstruction is performed using the proposed sub-pixel convolutional layer. We can directly reconstruct the image utilizing the convolutional features from the deep convolutional network. Lim et al. proposed an enhanced depth residual network (EDSR) [16], which made a significant performance through the deeper network. Other deep network like RDN [32] and MemNet [22], are based on dense blocks. Some networks focus on feature correlations in channel dimension, such as RCAN [31]and SAN [4].

The WDSR [28]proposed by Yu et al. draws two conclusions. First, when the parameters and calculations are the same, the model with more features before the activation function has better performance. Second, weight normalization (WN layer) can improve the accuracy of the network. In WDSR, there is a broader channel before the activation function of each residual block. Wang et al. proposed an adaptive weighted super-resolution network(AWSRN) based on WDSR [28]. It designs a local fusion block for more efficient residual learning. Besides, an adaptive weighted multi-scale model is developed. The model is used to reconstruct features and has superior performance in methods with roughly equal parameters.

Cao et al. proposed an improved Deep Residual Network (IDRN) [2]. It makes simple and effective modifications to the structure of residual blocks and skip-connections. Besides, a new energy-aware training loss EA-Loss was proposed. And it employs lightweight networks to achieve fast and accurate results. The SR feedback network (SRFBN) [15] proposed by Li et al. applies the RNN with constraints to process feedback information and performs feature reuse.

The Deep Plug and Play SR Network (DPSR)  [29] proposed by Zhang et al. can process LR images with arbitrary fuzzy kernels. Zhang et al. [30] obtained real sensor data by optical zoom for model training. Xu et al. [26]generated training data by simulating the digital camera imaging process. Their experiments have shown that SR using raw data helps to restore fine detail and clear structure.

Figure 1: The architecture of our proposed Adaptive Densely Connected Super-Resolution Reconstruction (ADCSR).The top is ADCSR,the middle are ADRU and AFSL,the bottom is ADRB.

3 Our Model

3.1 Network Architecture

As shown in Figure 1, our ADCSR mainly consists two parts: SKIP and BODY. The SKIP just uses the sub-pixel convolution [20]. The BODY includes multiple ADRUs (adaptive, dense residual units), GFF (global feature fusion layer) [32], and an AFSL layer(adaptive feature sub-pixel reconstruction layer). The model takes the RGB patches from the LR image as input. On the one hand, the HR image is reconstructed by SKIP using the low-frequency information of the LR images. On the other hand, the image is reconstructed by BODY using the high-frequency information of the LR images. We can obtain the final complete reconstructed HR image by combining the results of SKIP and BODY.

SKIP consists of a single or multiple sub-pixel convolutions with a convolution kernel size of 5. we have:


where represents the output of skip part, denotes the input image of LR and represents the sub-pixel convolution, which convolution kernel size is 5.

In the BODY, first, we use a convolution layer to extract the shallow features from the LR image.


where represents the feature extraction convolution, which kernel size is 3.

Second, we use several ADRUs to extract the deep features. There are four ADRBs (adaptive dense residual blocks) through adaptive dense connections in Each ADRU. The features are merged by the LFF (Local Feature Fusion Layer) and combined with a skip connection as the output of the ADRU. Each ADRB combines four convolution units by the same adaptive dense connection structure as ADRU. The convolution units adopt a convolution structure, which is similar to WDSR [28], including two layers of wide active convolution and one layer of Leakyrelu. After that, we fuse features by LFF, which combined with a skip connection as the output of the ADRB. GFF fuses the outputs of multiple ADRUs by means of concatenation and convolution.


where denotes the input feature map of th ADRU, means the function of th ADRU, , are hyperparameters.


means the output of the th ADRU, and represents the output of the last ADRU, which includes the skip connection.

The third part of BODY uses the GEF to combine all the output of ADRU, which fuses features by two convotion layers.


where means feature fusion.

Finaly, Image upsampling via AFSL. The AFSL consists of four sub-pixel convolution branches of different scales with a convolution kernel size of 3, 5, 7, and 9, respectively. The output is obtained through the junction layer and a single layer convolution.


In the second stage of BODY, the feature amplification layer is also implemented by a single convolution layer. The whole BODY is:


represents the output of BODY. The whole network can be expressed by formulas (8).


3.2 ADRB and ADRU

We will demonstrate the superiority of the adaptive dense connection structure in Chapter 4. To use as much adaptive residual structure as possible, we split the ADRU into ADRBs using adaptive dense connections and split ADRB into densely connected convolution units. At the same time, to get better results with less parameter amount, we use the residual block in WDSR as our convolution unit. As shown in Figure 1, ADRB and ADRU have similar connection structure. ADRB contains four convolution units, each of which can be represented by equotion (9).


where means the input of the convolution units. The kernel size of the is , and the is , is the input channels of the convolution units.

The whole ADRB can be expressed by equation (10).


where means convolution unit, denotes the input of ADRB. are hyperparameter, denotes the input of th convolution unit, represents the output of th convolution unit.

The whole ADRU can be formulated by equation (11).


3.3 Implementation

In this section, we will give specific implementation details. In SKIP, the convolution channel for the sub-pixel convolutional layer is defined as 5. The convolution kernel size of the LFF in BODY is 1.The two convolution kernel sizes of GFF are 1 and 3, respectively. In AFSL, the convolution kernels are 3, 5, 7 and 9. All other convolution kernel sizes are set to 3. There are 4 ADRUs in BODY. The number of output channels in feature extraction layer, convolution unit, LFF, and GFF are 128, and the 4 sub-pixel convolutions and the final output in AFSL are 3. The stride size is 1 throughout the network while using Leakyrelu as the activation function.

4 Experiments

4.1 Adaptive dense connections

We propose a structure for adaptive dense connections such as ADRU, and verify its performance through experiments. In the experiment, we designed three models. The model parameters are the same, and the calculations are roughly equal. The structure of the models is similar to the ADCSR consisting of a single ADRU. These three models are:
a. Add LFF [32] on WDSR [4] (to obtain the same model depth);
b. Add a dense connection based on a;
c. Add parameter adaptation based on b.
The three models have the same training parameters. We train our models with the data set DIV2K [16]. We also compare the performance on the standard benchmark dataset: B100 [17]. The number of iterations is 200. The learning rate is and halved at every 100 epochs. As shown in Figure 2, networks with dense connections and parameter adaptation have the highest performance under the same conditions.

Figure 2: Convergence analysis of tests on B100 with scaling factor during different model structures

4.2 Adaptive sub-pixel reconstruction layer (AFSL)

We test the reconstruction layer in BODY. We have designed a new reconstruction network model AFSL. To verify the performance of the model, we designed a straightforward model for comparison experiments. The model only includes the feature extraction layer and the reconstruction layer. As shown in Figure 3, the reconstruction layers are Sub-pixel convolution [20], AWMS [25], and AFSL. We performed the task on scale . The feature extraction layers and experimental parameters of the models are the same. We tested the models with B100 [17] and Urban100 [8]. At the same time, we also analyzed the difference in the number of FLOPs and model parameters. The result is shown in Table 1. We can see that AWMS and AFSL require more calculations and parameters than Sub-pixel convolution while its performance is better. In the case where the setting and the calculated amount are the same, the performance of AFSL is slightly better than AWMS.

Figure 3: Test model and structural Comparison of Three Reconstruction Layers
B100 Urban100 FLOPs Params
Sub-conv 30.402 27.750 0.02G 9K
AWMS 30.590 27.956 0.30G 128K
AFSL 30.592 27.958 0.30G 128K
Table 1: Performance comparison of three reconstruction layers

4.3 Pre-training SKIP

We have explored a training method that performs a separate pre-training of SKIP while training the entire model. This training method is used to make SKIP focus on the reconstruction of low-frequency information, while BODY focuses on high-frequency information reconstruction. We employ the same model, that is, the ADCSR containing a single ADRU with the same training parameters. But we train the model in different ways:
a. Train the entire network directly;
b. First pre-train SKIP, then train the whole network at the same time;
c. First pre-train SKIP, then set SKIP to be untrainable when training the entire network.

Figure 4 compares the image and image spectrum of SKIP and BODY output for models a and b. By comparing the output images, it can be seen that the BODY of the pre-trained SKIP model focuses on learning the texture edge details of the image. From the comparison of the output spectrum of the BODY part, the spectrogram of the pre-trained SKIP model is darker near the center and brighter around. It proves that the proposed method makes the BODY use more high-frequency information and less low-frequency information.

Figure 4: Results of pre-training SKIP on SKIP output

Figure 5 is a comparison of the test curves of the model on the B100 under different training modes. We found that networks that were pre-trained with SKIP achieved higher performance. And the network performance of tests b and c are similar.

Figure 5: Convergence analysis of tests on B100 with scaling factor during different model training methods

4.4 Training settings

We train our network with dataset DIV2K and Flickr2K [16]. The training set has a total of 3,450 images without data augmentation. DIV2K is composed of 800 images for training while 100 images each for testing and validation. Flickr2K has 2,650 training images. The input image block size is . SKIP is trained separately, and then the entire network is trained at the same time. The initial learning rate is . When the learning rate drops to , the training stops. we also adopt L1 loss to optimize our model. We train the network of scale firstly. Subsequently, when training the network of scale , , the BODY parameter of the scale is loaded (excluding the parameters of the AFSL). We train the model through the NVIDIA RTX2080Ti. Pytorch1.1.0+Cuda10.0+cudnn7.5.0 is selected as the deep learning environment.

method scale Set5 [1] Set14 [27] B100 [17] Urban100 [8] manga109 [18]
VDSR [10]
LapSRN [12]
MemNet [22]
EDSR [16]
RDN [32]
RCAN [31]
SAN [4]
33.66 0.9299 30.24 0.8688 29.56 0.8431 26.88 0.8403 30.80 0.9299
36.33 0.9542 32.45 0.9067 31.36 0.8879 29.50 0.8946 35.60 0.9663
37.53 0.9590 33.05 0.9130 31.90 0.8960 30.77 0.9140 37.22 0.9750
37.52 0.9591 33.08 0.9130 31.08 0.8950 30.41 0.9101 37.27 0.9740
37.78 0.9597 33.28 0.9142 32.08 0.8978 31.31 0.9195 37.72 0.9740
38.11 0.9602 33.92 0.9195 32.32 0.9013 32.93 0.9351 39.10 0.9773
38.24 0.9614 34.01 0.9212 32.34 0.9017 32.89 0.9353 39.18 0.9780
38.27 0.9614 34.12 0.9216 32.41 0.9027 33.34 0.9384 39.44 0.9786
38.31 0.9620 34.07 0.9213 32.42 0.9028 33.10 0.9370 39.32 0.9792
38.33 0.9619 34.48 0.9250 32.47 0.9033 33.61 0.9410 39.84 0.9798
38.38 0.9620 34.52 0.9252 32.50 0.9036 33.75 0.9418 39.97 0.9800
VDSR [10]
LapSRN [12]
MemNet [22]
EDSR [16]
RDN [32]
RCAN [31]
SAN [4]
30.39 0.8682 27.55 0.7742 27.21 0.7385 24.46 0.7349 26.95 0.8556
32.75 0.9090 29.30 0.8215 28.41 0.7863 26.24 0.7989 30.48 0.9117
33.67 0.9210 29.78 0.8320 28.83 0.7990 27.14 0.8290 32.01 0.9340
33.82 0.9227 29.87 0.8320 28.82 0.7980 27.07 0.8280 32.21 0.9350
34.09 0.9248 30.00 0.8350 28.96 0.8001 27.56 0.8376 32.51 0.9369
34.65 0.9280 30.52 0.8462 29.25 0.8093 28.80 0.8653 34.17 0.9403
34.71 0.9296 30.57 0.8468 29.26 0.8093 28.80 0.8653 34.13 0.9484
34.74 0.9255 30.65 0.8482 29.32 0.8111 29.09 0.8702 34.44 0.9499
34.75 0.9300 30.59 0.8476 29.33 0.8112 28.93 0.8671 34.30 0.9494
34.86 0.9305 30.81 0.8505 29.40 0.8127 29.44 0.8767 34.95 0.9521
34.93 0.9310 30.88 0.8514 29.43 0.8133 29.57 0.8784 35.11 0.9528
VDSR [10]
LapSRN [12]
MemNet [22]
EDSR [16]
RDN [32]
RCAN [31]
SAN [4]
28.42 0.8104 26.00 0.7027 25.96 0.6675 23.14 0.6577 24.89 0.7866
30.45 0.8628 27.50 0.7513 26.90 0.7101 24.52 0.7221 27.58 0.8555
31.35 0.8830 28.02 0.7680 27.29 0.7251 25.18 0.7540 28.83 0.8870
31.54 0.8850 28.19 0.7720 27.32 0.7270 25.21 0.7560 29.09 0.8900
31.74 0.8893 28.26 0.7723 27.40 0.7281 25.50 0.7630 29.42 0.8942
32.46 0.8968 28.80 0.7876 27.71 0.7420 26.64 0.8033 31.02 0.9148
32.47 0.8990 28.81 0.7871 27.72 0.7419 26.61 0.8028 31.00 0.9173
32.63 0.9002 28.87 0.7889 27.77 0.7436 26.82 0.8087 30.40 0.9082
32.64 0.9003 28.92 0.7888 27.78 0.7436 26.79 0.8068 31.18 0.9169
32.77 0.9013 29.02 0.7917 27.86 0.7457 27.15 0.8174 31.76 0.9212
32.82 0.9020 29.09 0.7930 27.90 0.7466 27.27 0.8197 31.98 0.9232
Table 2: Quantitative evaluation of competing methods. We report the performance of state-of-the-art algorithms on widely used publicly available datasets, in terms of PSNR (in dB) and SSIM. The best results are highlighted with read color while the blue color represents the second-best SR.
Figure 6: Visual results with bicubic degradation model() on Urban100
Figure 7: Two-stage adaptive dense connection super-resolution reconstruction network (DSSR)

4.5 Results with Bicubic Degradation

In order to verify the validity of the model , we compare the performance on five standard benchmark datasets: Set5 [1], Set14 [27], B100 [17], Urban100 [8], and manga109 [18]. In terms of PSNR, SSIM and visual effects, We compare our models with the state-of-the-art methods including Bicubic, SRCNN [5], VDSR [10], LapSRN [12], MemNet [22], EDSR [16], RDN [32], RCAN [31], SAN [4]. We also adopt self-ensemble strategy [16] to further improve our ADCSR and denote the self-ensembled ADCSR as ADCSR+. The results are shown in Table 2. As can be seen from the table, the PSNR and SSIM of the algorithm in , , exceed the current state of the art.

Figure 6 show the Qualitative comparison of our models with Bicubic, SRCNN [5], VDSR [10], LapSRN [12], MSLapSRN [13], EDSR [16], RCAN [31], and SAN [4] . The images of SRCNN, EDSR, and RCAN are derived from the author’s open-source model and code. Test images for VDSR, LapSRN, MSLapSRN, SAN are provided by their respective authors. In the comparison chart of img044 in Figure 6, the image reconstructed by the algorithm is clear and close to the original image. In img004, our algorithm has a better visual effect.

5 AIM2019: Extreme Super-Resolution Challenge

This work is initially proposed for the purpose of participating in the AIM2019 Extreme Super-Resolution Challenge. The goal of the contest is to super-resolve an input image to an output image with a magnification factor and the challenge is called extreme super-resolution.

Our model is the improved ADCSR, a two-stage adaptive dense connection super-resolution reconstruction network (DSSR). As shown in Figure 7, the DSSR consists of two parts, SKIP and BODY. The SKIP is a simple sub-pixel convolution [20]. The BODY part is divided into two stages. The first stage includes a feature extraction layer, multiple ADRUs (adaptive, dense residual units), GFF (global feature fusion layer) [32], and an AFSL layer (adaptive feature sub-pixel reconstruction layer).The second stage includes a feature amplification layer, an ADRB (adaptive dense residual block), and an AFSL.

During the training of DSSR, the network converges slowly due to the large network. We divide the network into two parts for training to speed up network convergence. When training DSSR, we first train the SKIP. The network ADCSR of scale is used as a pre-training parameter while training the entire network. At the same time, the feature extraction layer of the first level and each ADRU are set to be untrainable. During the period, GFF, AFSL and later second-level network parameters are trained at normal learning rates . Finally, we train the entire network when the learning rate is small. We train DSSR with dataset DIV8K. Other training settings are the same as ADCSR. Our model final result on the full resolution of the DIV8K test images is () : PSNR = 26.79, SSIM = 0.7289.

6 Conclusions

We propose an adaptive densely connected super-resolution reconstruction algorithm (ADCSR). The algorithm is divided into two parts: BODY and SKIP. BODY improves the utilization of convolution features by adaptively dense connections. We also explore an adaptive sub-pixel reconstruction layer (AFSL) to reconstruct the features of the BODY output. We pre-train SKIP in advance so that the BODY focuses on high-frequency feature learning. Several comparative experiments demonstrate the effectiveness of the proposed improved method. On the standard datasets, the comparisons of PSNR, SSIM, and visual effects show that the proposed algorithm is superior to the state-of-the-art algorithms.


  1. M. Bevilacqua, A. Roumy, C. Guillemot and M. L. Alberi-Morel (2012) Low-complexity single-image super-resolution based on nonnegative neighbor embedding. Cited by: §4.5, Table 2.
  2. Y. Cao, Z. He, Z. Ye, X. Li, Y. Cao and J. Yang (2019) Fast and accurate single image super-resolution via an energy-aware improved deep residual network. Signal Processing 162, pp. 115–125. Cited by: §2.
  3. H. Chang, D. Yeung and Y. Xiong (2004) Super-resolution through neighbor embedding. In Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004., Vol. 1, pp. I–I. Cited by: §2.
  4. T. Dai, J. Cai, Y. Zhang, S. Xia and L. Zhang (2019) Second-order attention network for single image super-resolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 11065–11074. Cited by: §1, §1, §2, §4.1, §4.5, §4.5, Table 2.
  5. C. Dong, C. C. Loy, K. He and X. Tang (2014) Learning a deep convolutional network for image super-resolution. In European conference on computer vision, pp. 184–199. Cited by: §1, §2, §4.5, §4.5, Table 2.
  6. K. He, X. Zhang, S. Ren and J. Sun (2016) Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778. Cited by: §1.
  7. G. Huang, Z. Liu, L. van der Maaten and K. Q. Weinberger (2017-07) Densely connected convolutional networks. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Cited by: §1.
  8. J. Huang, A. Singh and N. Ahuja (2015) Single image super-resolution from transformed self-exemplars. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 5197–5206. Cited by: §4.2, §4.5, Table 2.
  9. T. Karras, T. Aila, S. Laine and J. Lehtinen (2017) Progressive growing of gans for improved quality, stability, and variation. arXiv preprint arXiv:1710.10196. Cited by: §2.
  10. J. Kim, J. Kwon Lee and K. Mu Lee (2016) Accurate image super-resolution using very deep convolutional networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 1646–1654. Cited by: §1, §4.5, §4.5, Table 2.
  11. J. Kim, J. Kwon Lee and K. Mu Lee (2016) Deeply-recursive convolutional network for image super-resolution. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 1637–1645. Cited by: §1.
  12. W. Lai, J. Huang, N. Ahuja and M. Yang (2017) Deep laplacian pyramid networks for fast and accurate super-resolution. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 624–632. Cited by: §4.5, §4.5, Table 2.
  13. W. Lai, J. Huang, N. Ahuja and M. Yang (2018) Fast and accurate image super-resolution with deep laplacian pyramid networks. IEEE transactions on pattern analysis and machine intelligence. Cited by: §4.5.
  14. C. Ledig, L. Theis, F. Huszár, J. Caballero, A. Cunningham, A. Acosta, A. Aitken, A. Tejani, J. Totz and Z. Wang (2017) Photo-realistic single image super-resolution using a generative adversarial network. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 4681–4690. Cited by: §1.
  15. Z. Li, J. Yang, Z. Liu, X. Yang, G. Jeon and W. Wu (2019) Feedback network for image super-resolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 3867–3876. Cited by: §2.
  16. B. Lim, S. Son, H. Kim, S. Nah and K. Mu Lee (2017) Enhanced deep residual networks for single image super-resolution. In Proceedings of the IEEE conference on computer vision and pattern recognition workshops, pp. 136–144. Cited by: §1, §1, §2, §4.1, §4.4, §4.5, §4.5, Table 2.
  17. D. Martin, C. Fowlkes, D. Tal and J. Malik (2001) A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics. Cited by: §4.1, §4.2, §4.5, Table 2.
  18. Y. Matsui, K. Ito, Y. Aramaki, A. Fujimoto, T. Ogawa, T. Yamasaki and K. Aizawa (2017) Sketch-based manga retrieval using manga109 dataset. Multimedia Tools and Applications 76 (20), pp. 21811–21838. Cited by: §4.5, Table 2.
  19. S. Schulter, C. Leistner and H. Bischof (2015) Fast and accurate image upscaling with super-resolution forests. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 3791–3799. Cited by: §2.
  20. W. Shi, J. Caballero, F. Huszár, J. Totz, A. P. Aitken, R. Bishop, D. Rueckert and Z. Wang (2016) Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network. In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 1874–1883. Cited by: §2, §3.1, §4.2, §5.
  21. W. Shi, J. Caballero, C. Ledig, X. Zhuang, W. Bai, K. Bhatia, A. M. S. M. de Marvao, T. Dawes, D. O’Regan and D. Rueckert (2013) Cardiac image super-resolution with global correspondence using multi-atlas patchmatch. In International Conference on Medical Image Computing and Computer-Assisted Intervention, pp. 9–16. Cited by: §2.
  22. Y. Tai, J. Yang, X. Liu and C. Xu (2017) Memnet: a persistent memory network for image restoration. In Proceedings of the IEEE international conference on computer vision, pp. 4539–4547. Cited by: §2, §4.5, Table 2.
  23. R. Timofte, V. De Smet and L. Van Gool (2013) Anchored neighborhood regression for fast example-based super-resolution. In Proceedings of the IEEE international conference on computer vision, pp. 1920–1927. Cited by: §2.
  24. T. Tong, G. Li, X. Liu and Q. Gao (2017) Image super-resolution using dense skip connections. In Proceedings of the IEEE International Conference on Computer Vision, pp. 4799–4807. Cited by: §1.
  25. C. Wang, Z. Li and J. Shi (2019) Lightweight image super-resolution with adaptive weighted learning network. arXiv preprint arXiv:1904.02358. Cited by: §1, §4.2.
  26. X. Xu, Y. Ma and W. Sun (2019) Towards real scene super-resolution with raw images. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1723–1731. Cited by: §2.
  27. J. Yang, J. Wright, T. S. Huang and Y. Ma (2010) Image super-resolution via sparse representation. IEEE transactions on image processing 19 (11), pp. 2861–2873. Cited by: §2, §4.5, Table 2.
  28. J. Yu, Y. Fan, J. Yang, N. Xu, Z. Wang, X. Wang and T. Huang (2018) Wide activation for efficient and accurate image super-resolution. arXiv preprint arXiv:1808.08718. Cited by: §1, §2, §3.1.
  29. K. Zhang, W. Zuo and L. Zhang (2019) Deep plug-and-play super-resolution for arbitrary blur kernels. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1671–1681. Cited by: §2.
  30. X. Zhang, Q. Chen, R. Ng and V. Koltun (2019) Zoom to learn, learn to zoom. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 3762–3770. Cited by: §2.
  31. Y. Zhang, K. Li, K. Li, L. Wang, B. Zhong and Y. Fu (2018) Image super-resolution using very deep residual channel attention networks. In Proceedings of the European Conference on Computer Vision (ECCV), pp. 286–301. Cited by: §1, §2, §4.5, §4.5, Table 2.
  32. Y. Zhang, Y. Tian, Y. Kong, B. Zhong and Y. Fu (2018) Residual dense network for image super-resolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2472–2481. Cited by: §1, §1, §2, §3.1, §4.1, §4.5, Table 2, §5.
  33. W. W. Zou and P. C. Yuen (2011) Very low resolution face recognition problem. IEEE Transactions on image processing 21 (1), pp. 327–340. Cited by: §2.
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
Add comment
Loading ...
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description