other · Figshare · 2024

Sub-pixel convolution upsampling module.

Hao Wu (65943), Yanxi Liu (467386), Zhongyang Jin (17896517), Yuan Zhou (179597)

Cited by 0Open access

Abstract

<div><p>An optimization algorithm based on the LMPEC algorithm is proposed to rectify the nameplate image to address the problem that overexposure and underexposure of the nameplate image of electrical equipment will make subsequent nameplate recognition difficult. In the network structure, the PS-UNet++ network is based on the sub-pixel convolution upsampling module, and the UNet++ network is constructed as the feature extraction sub-network of the optimization algorithm to extract more detailed information from the model. Smooth L1 loss is substituted for L1 loss in the loss function to prevent model oscillation. In addition, to increase the robustness of the model, an improved method built on the multi-scale training method is applied. The experimental results indicate that, among all comparison algorithms, the optimized algorithm performs the best on the data set of electrical equipment nameplate exposure the experimenter generated. Compared to the original LMPEC algorithm, the SSIM, PSNR, and PI image evaluation indices are enhanced by 5.6%, 5.1%, and 7.96%, respectively.</p></div>

Cite this paper

APA 7

(65943), H. W., (467386), Y. L., (17896517), Z. J., & (179597), Y. Z. (2024). Sub-pixel convolution upsampling module. Figshare. https://doi.org/10.1371/journal.pone.0300792.g006

MLA 9

(65943), Hao Wu, et al. “Sub-pixel convolution upsampling module.” Figshare, 2024, https://doi.org/10.1371/journal.pone.0300792.g006.

Chicago

(65943), Hao Wu, Yanxi Liu (467386), Zhongyang Jin (17896517), and Yuan Zhou (179597). 2024. “Sub-pixel convolution upsampling module.” Figshare. https://doi.org/10.1371/journal.pone.0300792.g006.

Harvard

(65943), H.W. et al. (2024) ‘Sub-pixel convolution upsampling module’, Figshare. Available at: https://doi.org/10.1371/journal.pone.0300792.g006.

BibTeX

@misc{659432024pixel,
  title = {{Sub-pixel convolution upsampling module.}},
  author = {(65943), Hao Wu and (467386), Yanxi Liu and (17896517), Zhongyang Jin and (179597), Yuan Zhou},
  journal = {Figshare},
  year = {2024},
  doi = {10.1371/journal.pone.0300792.g006}
}

Made from the index record. Check names, capitalisation and page numbers against the paper. Need another style? AI for Research has over 10,000.