Baltimore Convention Center
Baltimore, Maryland, United States
14 - 18 April 2019
Conference 10988
Automatic Target Recognition XXIX
Monday - Thursday 15 - 18 April 2019
Conference
Committee
show | hide
Conference Chairs
Program Committee
Program Committee continued...
Monday 15 April Show All Abstracts
Welcome
Monday 15 April 2019
8:00 AM - 8:05 AM
Location: Conv. Ctr. 318

Conference Chairs: Riad I. Hammoud, BAE Systems FAST Labs (United States): Timothy L. Overman, Lockheed Martin Space Systems Co. (United States)
Session 1:
Keynote Session I
Monday 15 April 2019
8:05 AM - 9:00 AM
Location: Conv. Ctr. 318
Session Chairs:
Riad I. Hammoud, BAE Systems (United States) ;
Timothy L. Overman, Lockheed Martin Space Systems Co. (United States)
The hype, hope and promise of learning machines (Keynote Presentation)
Paper 10988-1
Author(s): Kevin L. Priddy, Air Force Research Lab. (United States)
Show Abstract
Panel Discussion: Machine Learning for Automatic Target Recognition (ML4ATR)
Monday 15 April 2019
9:00 AM - 10:30 AM
Location: Conv. Ctr. 318

Panel Moderators: Riad I. Hammoud, BAE Systems FAST Labs (United States); Timothy L. Overman, Lockheed Martin Space Systems Co. (United States); Andres Rodriguez, Intel AI (United States)

Panel Members:

Hanlin Tang, Intel AI (United States)

Todd Rovito, Air Force Research Lab. (United States)
Todd V. Rovito is a Senior Research Computer Scientist. Mr. Rovito works on remote sensing exploitation research, and is currently focusing on passive 3D Reconstruction and Reasoning and Deep Learning Object Detection from commercial space satellite systems. Mr. Rovito is from Cincinnati, Ohio, and works as a Technical Lead for the Decision Science Branch. Todd started in the Sensors Directorate in 2003 where he helped develop the first demonstration of a Wide Area Motion Imagery system with a real time down link known as AngelFire. AngelFire was deployed in support of Operating Iraqi Freedom where a Chief of Staff, USMC, stated “the Marines are now experiencing the most significant technical advance in real-time imagery we have seen in years”. Then Todd worked one year working a modernization effort at the AMOS site to upgrade hardware and software for the 24th largest telescope in the world along with four other USAF telescopes. Next Todd become the Technical lead of the Electro-Optical Exploitation Algorithms team where their research has focused on big data image processing and passive 3D Reconstruction and Reasoning. Most recently Todd has been leveraging commercial space systems to augment current Intelligence Surveillance Reconnaissance capability. As a Computer Scientist Todd aggressively uses open source software to develop exploitation systems quickly, generate the best value, and deliver the best performance to our warfighters.

Wesam Adel Sakla, Lawrence Livermore National Lab. (United States)
Wesam Sakla is a machine learning and computer vision research scientist at Lawrence Livermore National Laboratory (LLNL). He obtained his PhD in electrical engineering from Texas A&M University in 2009. His current research interests include the use of deep convolutional neural networks (CNNs) for classification, recognition, and detection/localization applications. Wesam currently leads several projects within the Global Security directorate at LLNL focused on applying state-of-the-art deep machine learning algorithms for problems of interest to national security.

Erik P. Blasch, Air Force Research Lab. (United States)
Dr. Erik Blasch is a program officer at United States Air Force Research Laboratory (AFRL) Air Force Office of Scientifc Research (AFOSR). His research focus is on information fusion, target tracking, and pattern recognition compiling 800+ scientific papers and book chapters, 25 patents, and numerous tutorials. His recent book is Multispectral Image Fusion and Colorization (SPIE, 2018). He is an Associate Fellow of AIAA, Fellow of IEEE, and Fellow of SPIE.

Bingcai Zhang, BAE Systems (United States)
Dr. Zhang is a BAE Systems Global Engineering Fellow. He joined BAE Systems in September 1995 right out of University of Wisconsin-Madison, where he earned his Ph.D. in engineering college and MS in computer science. His research interests are: (1) deep learning and 3D computer vision; and (2) photogrammetry and 3D mapping. Dr. Zhang has four inventions: (1) Embedded Photogrammetry; (2) Next Generation Automatic Terrain Extraction and Automatic Spatial Modeler; (3) Automatic Feature Extraction; and (4) DeepObject. In the last four years, Dr. Zhang has developed DeepObject, using deep learning to detect objects from images and 3D point clouds. DeepObject has three innovations: (1) simplicity learning; (2) rotation pattern match; and (3) double CNNs.

The Machine Learning for Automatic Target Recognition (ML4ATR) session, part of the ATR Conference at SPIE Defense + Commercial Sensing 2019, highlights the accomplishments to date and challenges ahead in designing and deploying deep learning and big data analytics algorithms, systems, and hardware for ATR. It provides a forum for researchers, practitioners, solution architects and program managers across all the widely varying disciplines of ATR involved in connecting, engaging, designing solutions, setting up requirements, testing, and evaluating to shape the future of this exciting field.

ML4ATR topics of interest include training deep learning based ATR with limited measured/real data, multi-modal satellite/hyperspectral/sonar/FMV Imagery analytics, graph analytic multi-sensory fusion, change detection, pattern-of-life analysis, adversarial learning, trust, and ethics.

This year ML4ATR hosts panelists from government labs, research institutions, and defense R&D companies. Each panelist gives a short keynote talk about their projects on machine learning for ATR. The chairs of this session encourage attendees from the SPIE Defense + Commercial Sensing 2019 community to engage in the discussions with the panel members.
Session 2:
Advanced Algorithms in ATR I
Monday 15 April 2019
11:00 AM - 12:30 PM
Location: Conv. Ctr. 318
Session Chair:
Abhijit Mahalanobis, Univ. of Central Florida (United States)
Spatio-temporal deep learning for on-the-move threat detection using full motion EO/IR sensors
Paper 10988-6
Author(s): John Pierre, physicsAI (United States)
Show Abstract
HySARNet – a hybrid machine learning approach to synthetic aperture radar automatic target recognition
Paper 10988-7
Author(s): Ryan J. Soldin, Douglas N. MacDonald, Roger Rouse, Timothy L. Overman, Lockheed Martin Space Systems Co. (United States)
Show Abstract
Shape-based ATR for wide-area processing of satellite imagery (Invited Paper)
Paper 10988-8
Author(s): Stephen P. DelMarco, Victor Tom, Helen Webb, William Snyder, Christopher Jarvis, David Fay, BAE Systems (United States)
Show Abstract
Discrimination of forests and man-made targets in SAR images based on spectrum analysis
Paper 10988-9
Author(s): Bin Zou, Weike Li, Harbin Institute of Technology (China); Yu Xin, Institute of Remote Sensing and Digital Earth (China); Lamei Zhang, Harbin Institute of Technology (China)
Show Abstract
Lunch Break 12:30 PM - 1:50 PM
Session 3:
Advances in Machine Learning for ATR I
Monday 15 April 2019
1:50 PM - 3:10 PM
Location: Conv. Ctr. 318
Session Chair:
Firooz A. Sadjadi, Lockheed Martin Corp. (United States)
Explainable automatic target recognition
Paper 10988-2
Author(s): Moses W. Chan, Nhat Nguyen, Sundip Desai, Lockheed Martin Space Systems Co. (United States)
Show Abstract
A comparison of target detection algorithms using DSIAC ATR algorithm development data set
Paper 10988-3
Author(s): Abhijit Mahalanobis, Univ. of Central Florida (United States)
Show Abstract
Fundamentals of target classification using deep learning
Paper 10988-4
Author(s): Irene Tanner, Abhijit Mahalanobis, Univ. of Central Florida (United States)
Show Abstract
Simple linear regression model based data clustering
Paper 10988-5
Author(s): Bingcheng Li, Lockheed Martin Systems Integration-Owego (United States)
Show Abstract
Session 4:
Advanced Algorithms in ATR II
Monday 15 April 2019
3:40 PM - 4:40 PM
Location: Conv. Ctr. 318
Session Chair:
Leon Cohen, Hunter College (United States)
Evaluation of synthetic IR environments for training deep neural nets
Paper 10988-10
Author(s): Mark Jeiran, Christopher L. Howell, U.S. Army Night Vision & Electronic Sensors Directorate (United States); Nader M. Namazi, Georges Nehmetallah, The Catholic Univ. of America (United States)
Show Abstract
Fast and robust detection of oil palm trees using high-resolution remote sensing images
Paper 10988-11
Author(s): Maocai Xia, Tsinghua Univ. (China); Weijia Li, Runmin Dong, Tsinghua University (China); Juepeng Zheng, Tongji University (China)
Show Abstract
Semantic segmentation based large-scale oil palm plantation detection using high-resolution satellite images
Paper 10988-12
Author(s): Runmin Dong, Weijia Li, Haohuan Fu, Maocai Xia, Juepeng Zheng, Le Yu, Tsinghua Univ. (China)
Show Abstract
Monday Plenary Session
Monday 15 April 2019
5:00 PM - 6:45 PM
Location: Conv. Ctr. 316

Plenary Session and Presentation details
The future of battlefield “things” (Plenary Presentation)
Paper 11002-301
Author(s): Philip Perconti, U.S. Army Research Lab. (United States)
Show Abstract
Is clinical virtual reality ready for primetime? (Plenary Presentation)
Paper 11002-300
Author(s): Albert Skip Rizzo, The Univ. of Southern California (United States)
Show Abstract
Tuesday 16 April Show All Abstracts
Tuesday Plenary Session
Tuesday 16 April 2019
7:45 AM - 9:00 AM
Location: Conv. Ctr. 316

7:45 am to 8:15 am
Morning Coffee Meet and Greet

8:15 am to 9:00 am
Plenary Presentation

Plenary Session and Presentation details
Mosaic warfare (Plenary Presentation)
Paper 11015-100
Author(s): Timothy P. Grayson, Defense Advanced Research Projects Agency (United States)
Show Abstract
Session 5:
Keynote Session II
Tuesday 16 April 2019
9:10 AM - 10:10 AM
Location: Conv. Ctr. 318
Session Chairs:
Riad I. Hammoud, BAE Systems (United States) ;
Timothy L. Overman, Lockheed Martin Space Systems Co. (United States)
Supporting real-time ATR from the cloud to the edge (Keynote Presentation)
Paper 10988-13
Author(s): Richard W. Linderman, U.S. Dept. of Defense (United States)
Show Abstract
Session 6:
Advanced Algorithms in ATR III
Tuesday 16 April 2019
11:00 AM - 12:00 PM
Location: Conv. Ctr. 318
Session Chair:
Leon Cohen, Hunter College (United States)
Design of adversarial targets: fooling deep ATR systems
Paper 10988-15
Author(s): Uche Osahor, Nasser Nasrabadi, West Virginia Univ. (United States)
Show Abstract
Comparing classifiers that exploit random subspace projections
Paper 10988-26
Author(s): Donald Waagen, David Gray, Jamie Gantert, Air Force Research Lab. (United States); Donald Hulsey, Dynetics, Inc. (United States)
Show Abstract
Radar target recognition using wavelet-based features extracted from compressively sensed signatures
Paper 10988-27
Author(s): Ismail I. Jouny, Lafayette College (United States)
Show Abstract
Lunch Break 12:00 PM - 1:10 PM
Session 7:
Advances in Machine Learning for ATR II
Tuesday 16 April 2019
1:10 PM - 3:20 PM
Location: Conv. Ctr. 318
Session Chair:
Vahid R. Riasati, Northrop Grumman Aerospace Systems (United States)
Learning detection-based target distance estimation from single camera images
Paper 10988-16
Author(s): Barıs Çaglar Ayyildiz, Ozan Yardimci, Yavuz Burak Eldeniz, Roketsan A.S. (Turkey)
Show Abstract
Interpretable deep learning for sonar automatic target recognition
Paper 10988-17
Author(s): Johnny Chen, Jason M. Trader, Jason E. Summers, Applied Research in Acoustics LLC (United States)
Show Abstract
On generalization of deep learning recognizers in radar and overhead imagery (Invited Paper)
Paper 10988-18
Author(s): George S. Goley, Adam R. Nolan, Scott Kangas, Etegent Technologies, Ltd. (United States)
Show Abstract
The Efficient and Robust Machine Learning (EMRL) Center
Paper 10988-19
Author(s): Erik Blasch, Todd Rovito, Vincent Velten, Air Force Research Lab. (United States); Robert Nowak, Univ. of Wisconsin-Madison (United States); Rebecca Willett, The Univ. of Chicago (United States); Eric Heim, Air Force Research Lab. (United States)
Show Abstract
Multisource deep learning for situation awareness
Paper 10988-20
Author(s): Erik Blasch, Peter Zulch, Air Force Research Lab. (United States); Zheng Liu, The Univ. of British Columbia Okanagan (Canada); Yufeng Zheng, Alcorn State Univ. (United States); Uttam Majumder, Todd Rovito, Alexandar Aved, Air Force Research Lab. (United States)
Show Abstract
Characterization of CNN classifier performance with respect to variation in optical contrast, using synthetic electro-optical data
Paper 10988-21
Author(s): Christopher Menart, Olga Mendoza-Schrock, Air Force Research Lab. (United States); Colin Leong, Univ. of Dayton (United States)
Show Abstract
Session 8:
Advanced Algorithms in ATR IV
Tuesday 16 April 2019
3:40 PM - 5:20 PM
Location: Conv. Ctr. 318
Session Chair:
Bing Li, Lockheed Martin Rotary and Mission Systems (United States)
Fast and accurate target detection in overhead imagery using double convolution neural networks (Invited Paper)
Paper 10988-22
Author(s): Bingcai Zhang, BAE Systems (United States)
Show Abstract
Neural network classification of degraded imagery using soft labels: towards human-level performance with “accurate” probabilistic outputs? (Invited Paper)
Paper 10988-23
Author(s): Jeffery R. Philson, Technology Service Corp. (United States)
Show Abstract
Probability distribution and propagation of clutter noise
Paper 10988-24
Author(s): Leon Cohen, Hunter College (United States)
Show Abstract
Physically-realizable adversarial examples for convolutional object detection algorithms
Paper 10988-25
Author(s): David R. Chambers, Harold A. Garza, Southwest Research Institute (United States)
Show Abstract
Wednesday 17 April Show All Abstracts
Session 9:
Keynote Session III
Wednesday 17 April 2019
8:00 AM - 9:00 AM
Location: Conv. Ctr. 318
Session Chairs:
Riad I. Hammoud, BAE Systems (United States) ;
Timothy L. Overman, Lockheed Martin Space Systems Co. (United States)
Information extraction from high-resolution commercial satellite imagery (Keynote Presentation)
Paper 10988-28
Author(s): Jacek Grodecki, DigitalGlobe, Inc. (United States)
Show Abstract
Session 10:
Advanced Algorithms in Remote Sensing I
Wednesday 17 April 2019
9:00 AM - 12:00 PM
Location: Conv. Ctr. 318
Session Chair:
Timothy L. Overman, Lockheed Martin Space Systems Co. (United States)
Transfer learning for ATR: comparing deep learning to other machine learning approaches
Paper 10988-29
Author(s): Samuel Rivera, Matrix Research Inc. (United States); Olga Mendoza-Schrock, Todd Rovito, Air Force Research Lab. (United States)
Show Abstract
Deep learning based super-resolution of aerial and satellite imagery
Paper 10988-43
Author(s): Asif Mehmood, Air Force Research Lab. (United States)
Show Abstract
The detection and azimuth information display of infrared moving targets (Invited Paper)
Paper 10988-31
Author(s): Lei Liu, Nanjing Univ. of Science and Technology (China)
Show Abstract
Evaluating confidence on deep learning decisions using stochastic methods
Paper 10988-32
Author(s): Ryan A. Elwell, U.S. Army Communications-Electronics Research Development and Engineering Command (United States)
Show Abstract
Generalization ability of region proposal networks for multispectral person detection
Paper 10988-33
Author(s): Michael Teutsch, HENSOLDT Optronics GmbH (Germany)
Show Abstract
Modeling sensor system performance for resolved targets with non-human exploitation (Invited Paper)
Paper 10988-41
Author(s): Timothy D. Ross, Matrix Research Inc. (United States); Jeffrey P. Duffy, Richard J. Thomas, Wesley A. Jones, Jordan Garcia, Infoscitex Corp. (United States); Hyatt B. Baker, Andy C. Rice, Air Force Research Lab. (United States)
Show Abstract
Lunch Break 12:00 PM - 1:00 PM
Workshop: Automatic Object Recognition at Scale on GBDX
Wednesday 17 April 2019
1:00 PM - 4:00 PM
Location: Conv. Ctr. 318

Moderator: Michael S. Foster, DigitalGlobe (United States)

In this workshop, you will learn about creating an end-to-end machine learning prototype for Automatic Object Recognition on DigitalGlobe’s Geospatial Big Data Platform, GBDX. Part one will entail an introduction to GBDX, delivering high-speed geospatial content, developer utilities, and scalable cloud processing accessible via established Application Programming Interfaces to enable an ecosystem of industry geospatial providers. Part two consists of a machine learning primer, focused on object detection, along with lessons learned and preliminary results from associated GBDX implementations.

Seating is limited. Please sign-up for the workshop in advance by contacting lilliand@spie.org.
Thursday 18 April Show All Abstracts
Session 11:
Keynote Session IV
Thursday 18 April 2019
8:00 AM - 9:00 AM
Location: Conv. Ctr. 318
Session Chair:
Michael Teutsch, HENSOLDT Optronics GmbH (Germany)
Deep learning of ATR using limited data (Keynote Presentation)
Paper 10988-34
Author(s): John Haddon, Lockheed Martin (United States); Christopher Aasted, Lockheed Martin Corp. (United States)
Show Abstract
Session 12:
Advanced Algorithms in Remote Sensing II
Thursday 18 April 2019
9:00 AM - 11:50 AM
Location: Conv. Ctr. 318
Session Chair:
Michael Teutsch, HENSOLDT Optronics GmbH (Germany)
Cross-spectral face recognition with image quality disparity using image fusion
Paper 10988-35
Author(s): Zhicheng Cao, Yuanming Zhao, Liaojun Pang, Heng Zhao, Xidian Univ. (China)
Show Abstract
Applying image processing techniques to security data: towards cyber target recognition
Paper 10988-36
Author(s): Jeremy Straub, North Dakota State Univ. (United States)
Show Abstract
Nighttime periocular recognition at long standoffs with deep learned features
Paper 10988-37
Author(s): Zhicheng Cao, Yuanming Zhao, Liaojun Pang, Weiqiang Zhao, Heng Zhao, Xidian Univ. (China)
Show Abstract
Development of infrared image training data sets using synthetic image generation software
Paper 10988-38
Author(s): Arthur Stout, FLIR Systems, Inc. (United States)
Show Abstract
Real-time LiDAR 3D object detection in an industrial vehicle
Paper 10988-39
Author(s): Tasmia Reza, Clemson Univ. (United States); Lucas Cagle, Mississippi State Univ. (United States); Pan Wei, Amazon.com, Inc. (United States); John E. Ball, Mississippi State Univ. (United States)
Show Abstract
iECO learned matched filters for automatic target recognition in synthetic mid-wave infrared imagery
Paper 10988-40
Author(s): Stanton R. Price, U.S. Army Engineer Research and Development Ctr. (United States); Steven R. Price, Mississippi College (United States); Carey D. Price, U.S. Army Engineer Research and Development Ctr. (United States)
Show Abstract
Research on image processing and intelligent recognition of space debris
Paper 10988-42
Author(s): Linghua Guo, China Academy of Space Technology (China)
Show Abstract
Award Announcements
Thursday 18 April 2019
11:50 AM - 11:55 AM
Location: Conv. Ctr. 318

2019 ATR Best Student Paper Awards and overall Best Paper Awards

This award is specific to papers in the Automatic Target Recognition conference 10988.

In order to be considered for these awards:
  • Presenter must make their oral presentation as scheduled
  • Manuscript must be submitted to SPIE no later than the week of 20 March 2019.


  • FOR STUDENTS: In addition to the above requirements, to be considered for the Best Student Paper Award:
  • Student must be the presenting author at the conference
  • Student must be the leading author of the manuscript
  • Student must send a message to the conference chairs identifying themselves as a student. Please include your Tracking Number and Paper Title.


  • Please send to: timothy.l.overman@lmco.com and riad.hammoud@baesystems.com

    A panel of experts headed by the ATR conference chairs will evaluate all the papers, both for quality and content. Attention will be given to 1) the innovation, clarity, and style of both the oral presentation at the conference and the manuscript submitted for publication, and 2) the importance of the work to the field of ATR. The winners will be recognized in person at the 2019 ATR conference. They will also be formally notified by email, and will receive a certificate of award.

    Sponsored by:

    Back to Top