article · 2019

Nonlinear Multi-scale Super-resolution Using Deep Learning

Kenneth Tran, Ashkan Panahi, Aniruddha Adiga, Wesam Sakla, Hamid Krim

Cited by 7Abstract and details only

Abstract

We propose a deep learning architecture capable of performing up to 8× single image super-resolution. Our architecture incorporates an adversarial component from the super-resolution generative adversarial networks (SRGANs) and a multi-scale learning component from the multiple scale super-resolution network (MSSRNet), which only together can recover smaller structures inherent in satellite images. To further enhance our performance, we integrate progressive growing and training to our network. This, aided by feed forwarding connections in the network to move along and enrich information from previous inputs, produces super-resolved images at scaling factors of 2, 4, and 8. To ensure and enhance the stability of GANs, we employ Wasserstein GANs (WGANs) during training. Experimentally, we find that our architecture can recover small objects in satellite images during super-resolution whereas previous methods cannot.

Cite this paper

APA 7

Tran, K., Panahi, A., Adiga, A., Sakla, W., & Krim, H. (2019). Nonlinear Multi-scale Super-resolution Using Deep Learning. https://doi.org/10.1109/icassp.2019.8682354

MLA 9

Tran, Kenneth, et al. “Nonlinear Multi-scale Super-resolution Using Deep Learning.” 2019, pp. 3182–86, https://doi.org/10.1109/icassp.2019.8682354.

Chicago

Tran, Kenneth, Ashkan Panahi, Aniruddha Adiga, Wesam Sakla, and Hamid Krim. 2019. “Nonlinear Multi-scale Super-resolution Using Deep Learning.” https://doi.org/10.1109/icassp.2019.8682354.

Harvard

Tran, K. et al. (2019) ‘Nonlinear Multi-scale Super-resolution Using Deep Learning’. Available at: https://doi.org/10.1109/icassp.2019.8682354.

BibTeX

@article{tran2019nonlinear,
  title = {{Nonlinear Multi-scale Super-resolution Using Deep Learning}},
  author = {Tran, Kenneth and Panahi, Ashkan and Adiga, Aniruddha and Sakla, Wesam and Krim, Hamid},
  pages = {3182--3186},
  year = {2019},
  doi = {10.1109/icassp.2019.8682354}
}

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