Topics of Interest
Deep learning methods have become an omnipresent and highly
successful part of recent approaches in imaging and
vision. However, in most cases they are used on a purely empirical
basis without real understanding of their behavior. From a scientific
viewpoint, this is unsatisfying.
Many mathematically inclined researchers have a strong desire to
understand the theoretical reasons for the success of these
approaches and to find relations between deep learning and
mathematically well-established techniques in imaging science.
The goal of this special issue is to showcase their latest research
results and to promote future research in this direction.
Topics of interest include, but are not limited to:
- gaining mathematical introspection into the behavior of
deep learning methods, e.g. through
- theoretical insights into their expressive power,
quality, stability, and efficiency
- analysis of their ability to handle the curse of
dimensionality
- investigation of their generalization properties
- theoretical bounds on their necessary complexity
- theories for architectural design
- characterization of their loss surface
- analysis of optimization algorithms
- mathematical theories for generative adversarial networks
- establishing connections between deep learning and
successful mathematical concepts in image analysis such as
- radial basis functions, splines, and harmonic analysis
- sparsity, compressed sensing, and dictionary learning
- subspace methods
- inverse problems, regularization theory, and operator learning
- variational methods, optimization, and optimal control
- ordinary and partial differential equations
- information theory, information geometry,
and the physics of information
- statistical learning theory
Gaining mathematical insights will be the decisive criterion for inclusion
into this special issue. Manuscripts which are primarily experimental are
not eligible.
Deadline and Submission Instructions
- Deadline for submission: December 31, 2018 (extended).
The printed special issue will appear in 2019. We aim at fast and
efficient reviewing, starting immediately after paper submission.
Since accepted manuscripts will become available direcly as
online first articles, earlier submission is helpful and encouraged.
If an individual manuscript requires substantially more time than
the others, it will be published in a regular JMIV issue.
- The usual
JMIV submission guidelines apply. All submissions will be
peer reviewed according to the JMIV standards.
Manuscripts that extend conference papers must contain at least
30 % novel material.
- Submissions must be uploaded through the regular login site
(Editorial Manager) of JMIV:
http://www.editorialmanager.com/jmiv/default.aspx
- Please make sure to choose the
"Mathematical Foundations of Deep Learning in Imaging Science"
Special Issue after logging in to the JMIV editorial manager.
This guarantees that your submission will be assigned to the
above guest editors.
For questions and more information, please contact
Joachim Weickert.
Imprint -
Data protection