SIAM J. Imaging Sciences
6 papers4 directions
SIAM J. Imaging Sciences is a publication venue focused on the mathematical and computational foundations of imaging. Within the machine learning and artificial intelligence landscape, it serves as a forum for research that bridges theoretical analysis with practical image processing applications. The venue is characterized by work on sparse signal recovery, including methods for overcoming shrinkage bias and addressing the gaps between theoretical guarantees and empirical performance. Research frequently explores high-dimensional asymptotics and finite-size effects in the context of image restoration and reconstruction.
The journal features contributions to multiscale and multiresolution analysis, specifically regarding anisotropic representations for capturing directional and geometric features. Methodologies published at this venue include universal shearlet systems for inpainting, dictionary learning for noisy or incomplete hyperspectral data, and the Split Bregman method for L1-regularized problems. Additionally, the venue hosts research on adaptive kernel-based methods, such as nonlocal means, which are often analyzed through the lens of oracle inequalities and minimax rates.
Prolific authors associated with this venue include Gitta Kutyniok, Joseph Salmon, and Charles-Alban Deledalle. Their research often addresses problems such as clustered sparsity, the separation of cartoon and texture components in images, and covariant least-square refitting. By focusing on the intersection of optimization, harmonic analysis, and statistical learning, the venue provides mathematical frameworks for computer vision and signal processing tasks.
Papers by Year
| 2017 | 1 |
| 2014 | 1 |
| 2013 | 1 |
| 2012 | 2 |
| 2009 | 1 |
Top Research Directions
Research directions most represented by papers from this venue.
Top Authors
| Author | Papers |
|---|---|
| Gitta Kutyniok | 2 |
| Joseph Salmon | 2 |
| Martin Genzel | 1 |
| Mingyuan Zhou | 1 |
| Nicolas Papadakis | 1 |
| Rebecca Willett | 1 |
| Samuel Vaiter | 1 |
| Tom Goldstein | 1 |