Anaheim Convention Center
Anaheim, California, United States
26 - 30 April 2020
Conference SI108
Algorithms for Synthetic Aperture Radar Imagery XXVII
Important
Dates
show | hide
Abstract Due:
16 October 2019

Author Notification:
20 December 2019

Manuscript Due Date:
1 April 2020

Conference
Committee
show | hide
Conference Chairs
Program Committee
  • Joshua N. Ash, Wright State Univ. (United States)
  • David Blacknell, Defence Science and Technology Lab. (United Kingdom)
  • Mujdat Cetin, Sabanci Univ. (Turkey)
  • Gil J. Ettinger, Systems & Technology Research (United States)
  • David A. Garren, Naval Postgraduate School (United States)
  • Eric R. Keydel, Leidos, Inc. (United States)
  • Juan Li, Univ. of Central Florida (United States)

Program Committee continued...
Additional Conference
Information
Z-FORMAT This conference follows a "Briefing, Poster Workshop, Panel Discussion" sequence known as the Z-format. During the first sessions of each day, authors highlight the results for their work in 10-minute oral briefings. After the presentations, these same authors are available for in-depth discussions in an extended poster session setting held in or near the conference room. Following the Poster Workshop, experts and audience address pressing issues and extensions from the sessions that day in a Panel Discussion.
Call for
Papers
Synthetic Aperture RADAR Research is advancing in several key application areas:
  • SAR target discrimination and classification algorithms and characterization of performance tradeoffs
  • moving target (vehicles, dismounts) detection, tracking, imaging, and classification exploiting the long integration times provided by SAR based MTI
  • video SAR for continuous surveillance
  • image compression for large area coverage and video SAR streams
  • ground, foliage, and building penetration
  • advanced detection algorithms including coherent and non-coherent change detection for finding difficult targets (e.g., targets deployed under tree cover, camouflage, etc.) and for discriminating decoys
  • 3D reconstruction and geolocation.
These enhancements are enabled by significant advancements in 2D and 3D imaging which are, in turn, driven by the incorporation of diversity into the imaging process. These diversities include: wide angle, polarization, waveform, frequency (e.g., Ka, Ku, X, L, UHF, VHF), and aperture (interferometric, MIMO, multi-static, passive sensing, and multi-pass sensing).

Of particular interest and importance is the application of machine learning (e.g., deep learning) approaches to these important problems. These very promising approaches are still in development and have the following challenges:
  • using machine learning with relatively small amounts of measured data for training including the generation and use of synthetic data
  • developing deep learning approaches that are robust, particularly when the conditions of the training and testing are mismatched
  • developing deep learning approaches that are self-aware of their performance (e.g., providing full posteriors conditioned on target, sensor, and environment states)
  • understanding the technical basis of a deep learning algorithm decision or estimate.
We strongly encourage papers to address these key challenges in applying machine learning to SAR applications and problems.

Context and Reproducibility

In order to provide context for technical contributions and enhance the reproducibility of results, authors are urged to explicitly characterize and state assumed models and model parameters/operating conditions affecting performance evaluations or simulations.

Challenge Problems

Previous conferences have revealed emerging needs for the following types of problems: compressive sensing, sparse aperture processing systems, change detection systems, foliage and building penetration systems, and adaptive ATR systems that adapt to changing conditions and requirements.
To facilitate the development of such systems, AFRL has published a number of challenge problems on the site: https://www.sdms.afrl.af.mil/

2020 Best Student Paper Award

In order to be considered for this award, the student must be the presenter and the primary author. A panel of experts will evaluate the papers, both for quality and content with regard to: 1) innovation, clarity, and style, and 2) the importance of the work to the field.
Back to Top