Huawei Yi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Huawei Yi has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Wenqian Xu
School of Electronics and Information Engineering, Liaoning University of Technology, Jinzhou 121001, China
2 shared publications - Xiaohui Li
School of Electronics and Information Engineering, Liaoning University of Technology, Jinzhou 121001, China
2 shared publications - Bo Li
School of Electronics and Information Engineering, Liaoning University of Technology, Jinzhou 121001, China
1 shared publication - Hang Qu
School of Electronics and Information Engineering, Liaoning University of Technology, Jinzhou 121001, China
1 shared publication - Huihui Qian
School of Electronics and Information Engineering, Liaoning University of Technology, Jinzhou 121001, China
1 shared publication - Jie Lan
College of Science, Liaoning University of Technology, Jinzhou 121001, China
1 shared publication
Latest Publications
- Open Access
DVDR-SRGAN: Differential Value Dense Residual Super-Resolution Generative Adversarial Network
In the field of single-image super-resolution reconstruction, GAN can obtain the image texture more in line with the human eye. However, during the reconstruction process, it is easy to generate artifacts, false textures, and large deviations in details between the reconstructed image and the Ground Truth. In order to further improve the visual quality, we study the feature correlation between adjacent layers and propose a differential value dense residual network to solve this problem. We first use the deconvolution layer to enlarge the features, then extract the features through the convolution layer, and finally make a difference between the features before being magnified and the features after being extracted so that the difference can better reflect the areas that need attention. In the process of extracting the differential value, using the dense residual connection method for each layer can make the magnified features more complete, so the differential value obtained is more accurate. Next, the joint loss function is introduced to fuse high-frequency information and low-frequency information, which improves the visual effect of the reconstructed image to a certain extent. The experimental results on Set5, Set14, BSD100, and Urban datasets show that our proposed DVDR-SRGAN model is improved in terms of PSNR, SSIM, and LPIPS compared with the Bicubic, SRGAN, ESRGAN, Beby-GAN, and SPSR models.
Open AccessDVDR-SRGAN: Differential Value Dense Residual Super-Resolution Generative Adversarial Network
In the field of single-image super-resolution reconstruction, GAN can obtain the image texture more in line with the human eye. However, during the reconstruction process, it is easy to generate artifacts, false textures, and large deviations in details between the reconstructed image and the Ground Truth. In order to further improve the visual quality, we study the feature correlation between adjacent layers and propose a differential value dense residual network to solve this problem. We first use the deconvolution layer to enlarge the features, then extract the features through the convolution layer, and finally make a difference between the features before being magnified and the features after being extracted so that the difference can better reflect the areas that need attention. In the process of extracting the differential value, using the dense residual connection method for each layer can make the magnified features more complete, so the differential value obtained is more accurate. Next, the joint loss function is introduced to fuse high-frequency information and low-frequency information, which improves the visual effect of the reconstructed image to a certain extent. The experimental results on Set5, Set14, BSD100, and Urban datasets show that our proposed DVDR-SRGAN model is improved in terms of PSNR, SSIM, and LPIPS compared with the Bicubic, SRGAN, ESRGAN, Beby-GAN, and SPSR models.
- Open Access
- 21 March 2023
- Published by MDPI AG in Electronics
- Vol. 12 (6) , 1477
A Graph Neural Network Social Recommendation Algorithm Integrating the Multi-Head Attention Mechanism
Collaborative filtering recommendation systems are facing the data sparsity problem associated with interaction data, and social recommendations introduce user social information to alleviate this problem. Existing social recommendation methods cannot express the user interaction interest and social influence deeply, which limits the recommendation performance of the system. To address this problem, in this paper we propose a graph neural network social recommendation algorithm integrating multi-head attention mechanism. First, based on the user-item interaction graph and social network graph, the graph neural network is used to learn the high-order relationship between users and items and deeply extract the latent features of users and items. In the process of learning user embedding vector representation based on the social network graph, the multi-head attention mechanism is introduced to increase the importance of friends with high influence. Then, we make rating predictions for the target users according to the learned user embedding vector representation and item embedding vector. The experimental results on the Epinions dataset show that the proposed method outperforms the existing methods in terms of both Recall and Normalized Discounted Cumulative Gain.
- 21 March 2023
- Published by MDPI AG in Electronics
- Vol. 12 (6) , 1477
Open AccessA Graph Neural Network Social Recommendation Algorithm Integrating the Multi-Head Attention Mechanism
Collaborative filtering recommendation systems are facing the data sparsity problem associated with interaction data, and social recommendations introduce user social information to alleviate this problem. Existing social recommendation methods cannot express the user interaction interest and social influence deeply, which limits the recommendation performance of the system. To address this problem, in this paper we propose a graph neural network social recommendation algorithm integrating multi-head attention mechanism. First, based on the user-item interaction graph and social network graph, the graph neural network is used to learn the high-order relationship between users and items and deeply extract the latent features of users and items. In the process of learning user embedding vector representation based on the social network graph, the multi-head attention mechanism is introduced to increase the importance of friends with high influence. Then, we make rating predictions for the target users according to the learned user embedding vector representation and item embedding vector. The experimental results on the Epinions dataset show that the proposed method outperforms the existing methods in terms of both Recall and Normalized Discounted Cumulative Gain.
- Open Access
Gram-GAN: Image Super-Resolution Based on Gram Matrix and Discriminator Perceptual Loss
The solution of a high-resolution (HR) image corresponding to a low-resolution (LR) image is not unique in most cases. However, single-LR–single-HR supervision is widely adopted in single-image super-resolution (SISR) tasks, which leads to inflexible inference logic of the model and poor generalization ability. To improve the flexibility of model inference, we constructed a novel form of supervision, except for the ground truth (GT). Specifically, considering the structural properties of natural images, we propose using extra supervision to focus on the textural similarity of the images. As textural similarity does not account for the position information of images, a Gram matrix was constructed to break the limitations of spatial position and focus on the textural information. Besides the use of traditional perceptual loss, we propose a discriminator perceptual loss based on the two-network architecture of generative adversarial networks (GAN). The difference between the discriminator features used in this loss and the traditional visual geometry group (VGG) features is that the discriminator features can describe the relevant information from the perspective of super-resolution. Quantitative and qualitative experiments were performed to demonstrate the effectiveness of the proposed method.
Open AccessGram-GAN: Image Super-Resolution Based on Gram Matrix and Discriminator Perceptual Loss
The solution of a high-resolution (HR) image corresponding to a low-resolution (LR) image is not unique in most cases. However, single-LR–single-HR supervision is widely adopted in single-image super-resolution (SISR) tasks, which leads to inflexible inference logic of the model and poor generalization ability. To improve the flexibility of model inference, we constructed a novel form of supervision, except for the ground truth (GT). Specifically, considering the structural properties of natural images, we propose using extra supervision to focus on the textural similarity of the images. As textural similarity does not account for the position information of images, a Gram matrix was constructed to break the limitations of spatial position and focus on the textural information. Besides the use of traditional perceptual loss, we propose a discriminator perceptual loss based on the two-network architecture of generative adversarial networks (GAN). The difference between the discriminator features used in this loss and the traditional visual geometry group (VGG) features is that the discriminator features can describe the relevant information from the perspective of super-resolution. Quantitative and qualitative experiments were performed to demonstrate the effectiveness of the proposed method.