Corpus ID: 221995432

Deep Image Reconstruction using Unregistered Measurements without Groundtruth

@article{Gan2020DeepIR,
  title={Deep Image Reconstruction using Unregistered Measurements without Groundtruth},
  author={W. Gan and Yu Sun and C. Eldeniz and Jiaming Liu and Hongyu An and Ulugbek S. Kamilov},
  journal={ArXiv},
  year={2020},
  volume={abs/2009.13986}
}
  • W. Gan, Yu Sun, +3 authors Ulugbek S. Kamilov
  • Published 2020
  • Computer Science, Engineering
  • ArXiv
  • One of the key limitations in conventional deep learning based image reconstruction is the need for registered pairs of training images containing a set of high-quality groundtruth images. This paper addresses this limitation by proposing a novel unsupervised deep registration-augmented reconstruction method (U-Dream) for training deep neural nets to reconstruct high-quality images by directly mapping pairs of unregistered and artifact-corrupted images. The ability of U-Dream to circumvent the… CONTINUE READING

    Figures and Tables from this paper.

    References

    SHOWING 1-10 OF 17 REFERENCES
    RARE: Image Reconstruction Using Deep Priors Learned Without Groundtruth
    • 13
    • PDF
    VoxelMorph: A Learning Framework for Deformable Medical Image Registration
    • 196
    • PDF
    Self-Supervised Physics-Based Deep Learning MRI Reconstruction Without Fully-Sampled Data
    • 19
    • PDF
    Convolutional Neural Networks for Inverse Problems in Imaging: A Review
    • 254
    • PDF
    Noise2Noise: Learning Image Restoration without Clean Data
    • 300
    • Highly Influential
    • PDF
    Noise2Void - Learning Denoising From Single Noisy Images
    • 138
    • Highly Influential
    • PDF
    U-Net: Convolutional Networks for Biomedical Image Segmentation
    • 16,816
    • PDF
    Enhanced Deep Residual Networks for Single Image Super-Resolution
    • 1,413
    • PDF