Comparison of classifier methods: a case study in handwritten digit recognition. Visual-patch-attention-aware Saliency Detection.
Simultaneous Hallucination and Recognition of Low-Resolution Faces Based on Singular Value Decomposition. The authors declare no conflicts of interest. However, even if our algorithm is not added symmetric padding, it can also achieve the same effect as that of ResNet with symmetric padding. The restored image of ResNet without symmetric padding has the distinct artificial stitching traces, and the stitching traces have been improved in that of the ResNet with the symmetric padding.
The images in second row, from left to right, are the enlarged images of the red box corresponding to the position in the first row. Figure 8 shows the enlarged denoised images, and the images from left to right in the first row are input noisy image, restored image of ResNet, restored image of ResNet with “dilation convolution ” + ” symmetric padding,” and restored image of FbResNet. The main reason is that the sparse feedbacks have been added to FbResNet model and can be used to smooth the artificial traces at the seam of patches. It can be seen that the images in second column of Figure 7 have obvious artificial stitching trace nevertheless, it is almost impossible to find the presence of artificial traces from the images in last column. The experiment setting of different models had been shown in Table 1. The meaning of the parameters on each layer in Figure 2 is similar to that in Figure 1. In order to compare the performance at the same configuration with Figure 1, Figure 2a is also improved to Figure 2b, which is called ResNet with dilated convolution. The network structure of (a) is same as that of ResNet except for the depth besides of the first layer and the last layer, only 4 building blocks are used in the reformed ResNet. Figure 2 shows two kinds of network structures reformed from ResNet and the network structure of DnCNN. The reformed ResNet is shown in Figure 2. Because our training set is small, for comparison on the same network scale, we reduce the depth of the ResNet and set it to 10. Nevertheless, small training samples can be easily constructed. In our opinion, very deep network architecture requires a huge training set, but in many computer vision tasks, a large number of training samples is not easy to be obtained. In order to verify the effectiveness of the proposed FbResNet, the comparison with the other network structures has been performed.
The main task of FbResNet is to estimate the residual information between the input degraded image and the output clean image. The second is connected to the last layer. The first is connected to the middle of the dilation convolution. In order to ensure that the estimated residual information does not deviate greatly, two forward feedbacks from the first layer have been added. By using the increasing dilated factors, the first-half layers can learn the residual information using an enlarged receptive field, and the latter half layers can refine the residual information using the decreasing dilation factors. The number behind each middle layer is the dilation factors, which is set to 1, 2, 3, 4, 4, 3, 2 and 1, respectively. Eight “Convolution + Batch Normalization + ReLU” blocks are in the middle layers. This layer has no “Batch Normal” and “ReLU,” in other words, the information produced by this layer is the original information after filtering the input image, then it is used to estimate the residual information by feeding back to the middle and the last layer. “Convolution” block is in the first layer. Inspired by the residual learning structure, we propose the deep residual network with sparse feedback loops for image restoration, and the structure of FbResNet is shown in Figure 1.
0 kommentar(er)
