The problem of removing noise from images has a long and rich history. Despite the vast amount of literature and the quality of current algorithms, it remains an open challenge. This is especially true when using photo, video and cinema cameras, where the push towards ever increasing resolution and dynamic range implies ever increasing demands on denoising performance. This minisymposium presents state of the art methods that cover representative, but in no way exhaustive, aspects in this field, from noise models to algorithm evaluation, from fast real-time in-camera techniques to powerful and computationally intensive methods for off-line processing.

A Tour of Denoising: Form, Function, and Applications

Peyman Milanfar (Google Research)

Camera Noise and Noise Perception in Motion Pictures

Tamara Seybold (Technische Universität München)

Benchmarking denoising algorithms with real photographs

Stefan Roth (Technische Universität Darmstadt )

How to Improve Your Denoising Result Without Changing Your Denoising Algorithm

Stacey Levine (Duquesne University)

Toward efficient and flexible CNN-based solutions for denoising in photography

Wangmeng Zuo (Harbin Institute of Technology)

Restoration of noisy and compressed video sequences

Toni Buades (Universitat de les Illes Balears)

Modeling and removal of correlated noise: towards effective approximate models

Milla Mäkinen (Tampere University of Technology)

High-Dimensional Mixture Models For Unsupervised Image Denoising

Julie Delon (Université Paris Descartes)

Organizers:

Marcelo Bertalmío (University Pompeu Fabra)

Stacey Levine (Duquesne University)

Keywords:

denoising, image enhancement, image reconstruction, machine learning, partial differential equation models, video