| Adaptive Sensing and Fusion of Multi-Sensor Data and Historical Information |
06-Nov-2009 |
20 pages |
| Authors:
Lawrence Carin; SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
 | Context plays an important role when performing underwater classification, and in this report we examine context from two perspectives. First, the classification of items within a single task is placed within the context of distinct concurrent or previous classification tasks (multiple distinct data collections). This is referred to as multi-task learning (MTL), and is implemented here in a statistical manner, using a simplified form of the Dirichlet process. In addition, ... |
|
| Multi-Sensor Information Integration and Automatic Understanding |
Nov-2008 |
43 pages |
| Authors:
Matthew Welborn; Austin Eliazar; SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
| This document is submitted to ONR as a final report for the Year III effort of the ONR C2CS research carried out by the team of Signal Innovations Group (SIG), Lockheed Martin and NAVAIR (henceforth referred to as the research team) to developed technology to process general video data of interest for base and port security. This research effort has also produced a real-time implementation of the tracking and anomalous ... |
|
| Fast Electromagnetic Solvers for Large-Scale Naval Scattering Problems |
27-Sep-2008 |
17 pages |
| Authors:
Lawrence Carin; SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
| Efficient modeling of electromagnetic scattering has always been an active topic in the field of computational electromagnetics. To reduce the memory and CPU time in the method of moments (MoM) solution, an efficient method based on pseudo skeleton approximation is presented in this report. The algorithm is purely algebraic, and therefore its performance is not associated with the kernel functions in the integral equations. The algorithm starts with a multilevel ... |
|
| Classifier Design for Multi-aspect Low-Frequency Broadband Target Signatures |
25 JUN 2008 |
14 pages |
| Authors:
Lawrence Carin; Patrick Rabenold; SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
| The early classification algorithms developed for the Low Frequency Broadband (LFBB) mine identification program were implemented using signal features that treat each target aspect independently. This approach results in a trained classifier that ignores any spatial correlation between sequential target responses. This motivated the design of a correlation-based kernel that combines the advantages of a discriminative approach and the sequential nature of the data. The resulting correlation kernel directly evaluates ... |
|
| Optimal Sensor Management for Next-Generation EMI Systems |
13-Jun-2008 |
30 pages |
| Authors:
Lawrence Carin; Nilanjan Dasgupta; Hui Li; SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
| This document serves as the final report on the project titled Optimal Sensor Management for Next-Generation EMI Systems (SERDP Project MM-1591). This project is a collaboration between SIG, Dr. T. C. Bell of AETC, and Dr. Herb Nelson of NRL. This research is directed toward developing the adaptive sensor-management architecture needed for next-generation electromagnetic induction (EMI) systems. Specifically, SERDP and ESTCP are currently funding multi-coil EMI systems that provide significant ... |
|
| The Use of SIG Algorithms in Sea Tests |
28 NOV 2007 |
12 pages |
| Authors:
Lawrence Carin; Patrick Rabenold; SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
| This report summarizes recent progress by Signal Innovations Group (SIG) in supporting the Naval Research Laboratory (NRL) on the development and application of mine identification algorithms using a Low Frequency Broadband (LFBB) sonar system. SIG has the tasks of developing the algorithms and transitioning them to NRL for use in sea tests. Fully-transitioned algorithms have performed well in sea tests, with additional development increasing the computational efficiency. The discussion below ... |
|
| Multi-Sensor Information Integration and Automatic Understanding |
27 NOV 2007 |
8 pages |
| Authors:
Matthew Welborn; Samantha Venters; SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
| The purpose of this program is to address the development of algorithms for adaptive processing of multi-sensor data, employing feedback to optimize the linkage between observed data and sensor control. The envisioned multi-modal adaptive system is applicable for intelligence, surveillance, and reconnaissance (ISR) in general environments, addressing base and port security, as well as urban and suburban sensing during wartime and peacekeeping operations. Of significant importance for current and anticipated ... |
|
| Multi-Sensor Information Integration and Automatic Understanding |
AUG 2007 |
6 pages |
| Authors:
Matthew Welborn; Samantha Venters; SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
| During the last reporting period, important progress has been made in the system for both the object tracking and anomalous behavior algorithms. We have completed significant enhancements to the background color modeling that improves both the efficiency and performance of the overall system, allowing more consistent and robust object tracking and anomalous behavior detection. In addition, we have completed several data collection programs, including efforts that will support analysis for ... |
|
| Progress Report on Advanced Detection and Classification Algorithms for Acoustic-Color-Based Sonar Systems |
29 NOV 2006 |
15 pages |
| Authors:
Lawrence Carin; Nilanjan Dasgupta; Steven Haron; SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
| This report summarizes recent progress by Signal Innovations Group (SIG) in supporting the Naval Research Laboratory (NRL) on development of a new low-frequency sonar system. SIG has the tasks of developing the algorithms and transitioning them to NRL, for use in sea tests. The discussion below provides a summary of the following items: (i) a kernel-based matching pursuits classification algorithm, (ii) life-long learning, (iii) in situ learning, and (iv) a ... |
|
| Multi Sensor Information Integration and Automatic Understanding |
22 AUG 2006 |
29 pages |
| Authors:
SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
| This program addresses Automatic Image Understanding and Automatic Integration of Disparate Sources of Information. The techniques are particularly focused on asymmetric warfare, urban warfare, guerrilla warfare, and port/base security, for which automatic integration of disparate sources is particularly important, typically with very limited if any 'a priori' training data. Concerning automatic image understanding, we are principally considering image sequences (video). The approaches utilize the new field of semi-supervised learning. Specifically, ... |
|
| Advanced Detection and Classification Algorithms for Acoustic-Color-Based Sonar Systems |
29 JUN 2006 |
7 pages |
| Authors:
Lawrence Carin; SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
| Signal Innovations Group (SIG) has been working closely with the Naval Research Laboratory (NRL) on development of advanced algorithms for detection and classifying MCM targets, with data collected using the NRL sonar system. Over the current period of performance SIG has delivered to NRL kernel matching pursuits (KMP) software, that was employed by NRL at the most recent blind test. Details on the KMP algorithm are provided below. Additionally, NRL ... |
|
| Quarterly Report for 2/25/2006-05/24/2006 |
25 MAY 2006 |
6 pages |
| Authors:
SIGNAL INNOVATIONS GROUP INC DURHAM NC
|
| This research effort is focusing on development of adaptive sensing algorithms for asymmetric threats. The algorithms are particularly targeted towards use with data from multi-camera video surveillance systems. We are not attempting to model the infinite class of asymmetric targets, since these are generally unpredictable and therefore limited if any a priori sensor data are available for these threats. Rather, we model normal or typical behavior using statistical algorithms. Key ... |
|
Reports by Corporate Author
