Corpus ID: 167217647

Invertible generative models for inverse problems: mitigating representation error and dataset bias

@article{Asim2019InvertibleGM,
  title={Invertible generative models for inverse problems: mitigating representation error and dataset bias},
  author={Muhammad Asim and Ali Ahmed and Paul Hand},
  journal={ArXiv},
  year={2019},
  volume={abs/1905.11672}
}
  • Muhammad Asim, Ali Ahmed, Paul Hand
  • Published in ArXiv 2019
  • Computer Science
  • Trained generative models have shown remarkable performance as priors for inverse problems in imaging. For example, Generative Adversarial Network priors permit recovery of test images from 5-10x fewer measurements than sparsity priors. Unfortunately, these models may be unable to represent any particular image because of architectural choices, mode collapse, and bias in the training dataset. In this paper, we demonstrate that invertible neural networks, which have zero representation error by… CONTINUE READING

    Create an AI-powered research feed to stay up to date with new papers like this posted to ArXiv

    Citations

    Publications citing this paper.

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 24 REFERENCES

    Glow: Generative Flow with Invertible 1x1 Convolutions

    VIEW 10 EXCERPTS
    HIGHLY INFLUENTIAL

    Compressed Sensing using Generative Models

    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL

    Automated Flower Classification over a Large Number of Classes

    VIEW 6 EXCERPTS
    HIGHLY INFLUENTIAL

    Analyzing Inverse Problems with Invertible Neural Networks

    VIEW 1 EXCERPT