SPARS 2015

Signal Processing with Adaptive Sparse Structured Representations

July 6-9, 2015, Cambridge, UK.

Welcome to the SPARS 2015 website!

SPARS 2015 has now happened and approximately 150 delegates attended. The organisers would like to thank all the authors and speakers who contributed to its success.

The Technical Program is available here and slides of most of the talks are available for download at the end of the Program page (from 2 Sept 2015). Slides of Plenary talks are downloadable from the Plenary Talks page.

The Best Student Paper Prize was awarded to two very deserving student authors:

  • Ju Sun of Columbia University, USA, for paper 1111,
    'Complete Dictionary Recovery over the Sphere'
  • Yanjun Li of University of Illinois at Urbana Champaign, USA, for paper 1047,
    'Identifiability of blind deconvolution with subspace or sparsity constraints'.

SPARS 2017: The next SPARS will be held in Lisbon in June 2017. We hope to see you there. For information, please see the website:


The Signal Processing with Adaptive Sparse Structured Representations SPARS workshop seeks novel ideas and results, both experimental and theoretical, for adaptive sparse representations, sampling and computational methods for high-dimensional data that feature structured combinatorial and geometric foundations.

SPARS 2015 will take place at Robinson College, Cambridge, from July 6th to July 9th, 2015 and we welcome you! The conference banquet will take place in the 400-year-old Great Hall of Trinity College.

Important Dates

  • Submission of Paper Abstracts: January 21, 2015 (extended by 24 hours to avoid any problems with web congestion at midnight on 20th - London time)
  • Notification of Acceptance of Papers: April 22, 2015 (Delayed to April 29 - apologies)
  • Workshop Dates: July 6th to July 9th, 2015

Topics of interest:

  • Sparse coding and representations, vector quantization, and dictionary learning.
  • Sparse and low-rank approximation algorithms: performance and complexity analysis, new methodologies.
  • Compressive sensing and learning: new theory and methods.
  • Dimensionality reduction, feature extraction, classification, detection, and source separation.
  • Sparsity measures in approximation theory, information theory and statistics.
  • Regularization theory with low-complexity / low-dimensional structures.
  • Statistical models and algorithms for sparsity, including Bayesian, likelihood-based, entropy and variational Bayes.
  • Sparse network theory and analysis, including dynamic (time-varying) networks and large networks.
  • Applications of sparsity and low-rank ideas to areas such as 'recovery without phase' and inverse covariance estimation.
  • Big data applications, including but not limited to geophysics, neuroscience, biomedical imaging, array processing, genetics, optics and radar, and feedback control.

Plenary Speakers

We are delighted that the following highly respected researchers have agreed to give plenary talks at SPARS 2015:

  • Maryam Fazel
  • Zoubin Ghahramani
  • Venkat Chandrasekaran
  • Michael Lustig
  • Gabriel Peyre
  • Justin Romberg
  • Roman Vershynin
  • Martin Wainwright
Our Special Lecturer will be:

Financial Support

This workshop has been partly funded by the Raymond and Beverly Sackler Fund, which is administered by Trinity College, Cambridge, and also by the Conferences Fund of Trinity College. The organizers are most grateful for this support.

Previous SPARS events

SPARS 2015 will be the sixth in a series of workshops on sparse representation of signals, held approximately once every two years since 2005, whose websites may be reached via the following links: