Recent Advances in Algorithms for Automatic Target Detection and Recognition
Dr. Abhijit Mahalanobis of Lockheed Martin Corporation
Wednesday, March 28, 2018 · 2:30PM · HEC 101
Abstract
The problem of detecting and recognizing targets continues to be substantial interest to the DoD community for tactical and surveillance imaging systems.
Although many techniques have been developed over the years, challenges remain due to the limited number of pixels on targets (POT), lack of training images,
and dense background clutter. In this talk, we will fist review classical approaches for automatic target detection and recognition (ATD/R) based on
correlation filtering techniques that were optimized to handle such performance challenges. These include early results obtained with well-known correlation
filters such as the maximum average correlation height (MACH) filter and quadratic correlation filters (QCFs). We will then review methods for improving the
performance of correlation filters using machine learning techniques such as manifold analysis and support vector machines, as well as "multi-look" and
"collaborative" approaches that exploit sensor geometry and temporal information to enhance ATD/R performance. Most recently, Deep Learning (DL) methods
have shown significant promise for achieving very accurate detection and classification rates. We will review some recent results obtained using public
domain data sets, and discuss the benefits and challenges of DL algorithms in the context of ATD/R applications.
Biography
Dr. Abhijit Mahalanobis is a Senior Fellow of the Lockheed Martin Corporation. His primary research areas are in Systems for Information processing,
Computational Sensing and Imaging, and Video/Image processing for information exploitation and ATR. He has over 170 journal and conference publications in
this area. He also holds four patents, co-authored a book on pattern recognition, contributed several book chapters, and edited special issues of several
journals. Abhijit completed his B.S. degree with Honors at the University of California, Santa Barbara in 1984. He then joined the Carnegie Mellon
University and received the MS. and Ph.D. degrees in 1985 and 1987, respectively. Prior to joining Lockheed Martin, Abhijit worked at Raytheon in Tucson,
and was a faculty at the University of Arizona and the University of Maryland.
Abhijit was elected a Fellow of SPIE in 1997, and a Fellow of OSA 2004 for his work on optical pattern recognition and automatic target recognition. He was
elected Fellow of IEEE in 2015 for his work on the theory of correlation filters. He served as an associate editor for Applied Optics from 2004-2009. He
was as an associate editor for the journal of the Pattern Recognition Society from 1994-2003. He served on OSA's Science and Engineering council in the
capacity of Pattern Recognition Chair from 2001-2004, and as Technical Group Chair for Information Acquisition, Processing and Display on OSA's Board of
Meetings from 2012-2015. He also serves on the organizing committees for the SPIE conferences, and OSA's annual and topical meetings. Abhijit received the
Hughes Business unit Patent Award in 1998. He was recognized as the Innovator of the Year by the State of Arizona in 1999, and was elected to the Raytheon
Honors program for distinguished technical contribution and leadership. At Lockheed Martin, he was elected to the rank of Distinguished Member of Technical
Staff in 2000, and twice received the Lockheed Martin Technical Excellence award, the Author of the Year award in 2001, and the Inventor of the Year in 2005
for designing novel target recognition systems. In October 2005, he received the prestigious Lockheed Martin NOVA award, the Corporation's highest honor,
for putting together a National Team and a winning strategy in the FCS competition. Abhijit was also recognized as the 2006 Scientist of the Year by Science
Spectrum Magazine, a publication of the Career Communication Group, Inc.
