TITLE:

Image Restoration Using Learned Patch Priors

DATE:

Friday, January 25th, 2019

TIME:

3:30 PM

LOCATION:

GMCS-314

SPEAKER:

Shibin Parameswaran (SPAWAR)

ABSTRACT:

Cameras have become ubiquitous leading to an increase in the amount of video
and image data captured by amateurs and professionals alike. Their ease of
deployability makes them a great sensor for security applications as well.
Hence, there is an ever-growing need to *efficiently* process and enhance
captured image and videos for improving the performance of subsequent
computer vision algorithms or simply for aesthetic reasons. In this talk,
we present efficient techniques for large scale image and video denoising
by using learned patch priors. First, we present an efficient denoising
algorithm that uses a Gaussian Mixture Model (GMM) to model patch prior
for image restoration. It is two orders of magnitude faster than similar
methods while providing a very competitive quality-vs-speed operating
curve. Following this, we introduce an improved model for image patches
by proposing a more expressive distribution than GMM called Generalized
Gaussian Mixture Models (GGMM). We circumvent the prohibitive computational
complexity of using GGMM patch priors for image restoration by introducing
asymptotically accurate but computationally efficient approximations to
the bottlenecks encountered in this formulation. Our evaluations indicate
that the GGMM prior is consistently a better fit for modeling image patch
distribution and performs better on average in image denoising.

HOST:

Dr. Jerome Gilles, Department of Mathematics and Statistics

DOWNLOAD:

PDF File