Timezone: »

AGEM: Solving Linear Inverse Problems via Deep Priors and Sampling
Bichuan Guo · Yuxing Han · Jiangtao Wen

Tue Dec 10:45 AM -- 12:45 PM PST @ East Exhibition Hall B + C #107

In this paper we propose to use a denoising autoencoder (DAE) prior to simultaneously solve a linear inverse problem and estimate its noise parameter. Existing DAE-based methods estimate the noise parameter empirically or treat it as a tunable hyper-parameter. We instead propose autoencoder guided EM, a probabilistically sound framework that performs Bayesian inference with intractable deep priors. We show that efficient posterior sampling from the DAE can be achieved via Metropolis-Hastings, which allows the Monte Carlo EM algorithm to be used. We demonstrate competitive results for signal denoising, image deblurring and image devignetting. Our method is an example of combining the representation power of deep learning with uncertainty quantification from Bayesian statistics.

Author Information

Bichuan Guo (Tsinghua University)
Yuxing Han (South China Agriculture University)
Jiangtao Wen (Tsinghua University)