Seedling Information

This additional information has been shown theoretically to speed up the rate of convergence for learning. In a recent study, LUPI-based Support Vector Machines (SVM+) were found to achieve the same or better accuracy as an SVM model with a significant reduction in the amount of training data required. One challenge that this approach faces is that it requires 10 times the computation time as an SVM, which could reduce the practicality of the approach for many applications. The goal of the seedling is to reduce computation time to SVM levels without eliminating the benefits of the SVM+ approach and to further explore the types of privileged information that could benefit the Intelligence Community.

Performer: Applied Communication Sciences

As part of this effort, the performer (SRI International) will (1) study key needs, specifications, high-risk areas and potential approaches to building an end-to-end ARC3 system; (2) empirically study the implications of a dynamic information aperture with respect to trade-offs between efficiency, usability and reliability when automated, versus user-driven approaches are followed. This will be accomplished by developing experimental prototype interfaces for AR collaboration and corroboration between an analyst and a collector; and (3) empirically study techniques for automated and user-driven information extraction in the presence of uncertain and unreliable visual localization and scene understanding technologies. An experimental prototype will build user scenarios in which the interplay between audio-visual-text processing and user interaction can be systematically evaluated.

Performer: Johns Hopkins University Applied Physics Laboratory, SRI International

This effort increases the understanding of the limitations of latent recognition systems and source(s) of latent recognition errors. Such analysis and studies are difficult to carry out with COTS latent recognition systems because of their proprietary nature. This research will leverage advances in deep learning, multiscale feature extraction, and high-order graph based minutiae matching.

Performer: Michigan State University

A unique aspect of this work is the utilization of existing publicly-tagged data to create a hierarchical tree model of a region structured by city, neighborhood, and block and leverage that information for search and retrieval. This seedling’s results support the Finder program.

Performer: University of Central Florida

The researchers plan to quantitatively analyze the sources of failure in current acoustic modeling approaches. The focus of the ASR research community has been on improving performance with increasing amounts of matched-condition training data and language modeling. Using a diagnostic approach, their prior research has revealed that many errors are due to mismatch between the data and the statistical assumptions built into the acoustic models. A key idea in the current effort is to use resampling at different levels to systematically modify the data in order to selectively match particular assumptions in the model. This will be used to discern specific error conditions and the extent of their contribution to overall error rate as well as providing a quantitative metric for comparing data sets with regard to factors affecting performance. An important aspect of this research is training/test mismatch (a condition in which current ASR systems fail quite badly). In addition, the research team will ask experts in the field where they believe ASR technology is effective, where it fails, and what its shortcomings are. This survey will include both interviews and a review of the literature. This research effort will produce new quantitative methods and metrics for ASR acoustic model failure analysis, a report on failure modes and lessons learned, and suggestions for future research. The software developed in the research will be made freely available to the research community.

Performer: SRI International

Resulting Research: FA8650-15-C-9101

2) Define quantitative metrics and qualitative mechanisms for assessing and improving the quality of automatically built models, 3) Idenfity the problem classes, data types, and model families for which automating model discovery is likely to be feasible, and 4) Demonstrate infrastructure that makes it cost-effective to integrate automatic and manual model discovery approaches.

Performer: Massachusetts Institute of Technology

Resulting Research: 2015-15061000003

The goal is to use the models to examine under what conditions groups are more likely to succumb to "cognitive collapse," and what interventions if any might be used to steer groups away from this undesirable state. "Cognitive collapse" refers to a state in which a group becomes fixated on a single hypothesis, failing to take into account contradictory evidence or to consider alternative hypotheses. The seedling completed at the end of 2QFY2011, with results demonstrating that the simulation-based tools provide a viable approach to exploring this type of group behaviors.

Performer: Jacobs Strategic Solutions Group, Inc.

The proposed Data Engine has three key features. First, the engine will store and manage uncertain data, so uncertainty will be an integral part of its data model and algorithms. Second, the engine will be general purpose, meaning that the engine can be used in a variety of workbench settings, by many different inferencing algorithms, and by different applications. Thus, the engine will support a data model, query language, and interface that are appropriate for a wide spectrum of needs. Third, the engine will be high performance and it will be able to efficiently process vast amounts of data, often residing on disk or slow memory. The user will have the option of accepting approximate results in order to improve running times further.
Language Computer Corporation is exploring the automatic identification and characterization of emerging events as reported in textual information feeds, such as news reports. Complex events of the sort of interest to the IC are made up of multiple component events and contextual influences which over time can converge into an action or event of interest. By identifying component events and tracking them over time, event structures can emerge, supporting indicators and warnings to analysts as an input to their task prioritization process.

Performer: Stanford University

Resulting Research: N10PC20259

Because of its compelling implications (no glasses required) to the future of 3D displays, Innovation Magazine identified this technology in the top three developments of 2010, right behind the iPad. The seedling permitted characterization of performance requirements for holographic displays allowing collaborative analysis with no head-gear, and the quantification of many metrics that formed the Synthetic Holographic Observation program. This is a good example of a seedling being used in the creation of a program.

Performer: Sandia National Lab

Resulting Research: N/A

Indeed, the results showed that, after the system learned the depth cues from realistic 3D imagery, performance increased for 2D imagery where the system successfully applied the 3D depth cues absent in the 2D images. These insights may guide the development of significantly more sophisticated automated image analysis systems.

Performer: eCortex, Inc.

Performer: Carnegie Mellon University

This research builds on a body of research that suggests there are indicators of behavioral and even physiological synchrony that emerge during such interactions and can provide clues to the quality of the exchange and eventual outcomes. Application of this research remains limited because it has primarily used subjective measures of behavior, relied on methods that have low temporal precision, and often treats each member of a dyad as an individual, rather than analyzing the interaction as a pair. Draper will seek to advance this research by applying linear and non-linear methods to analyze physiological and non-verbal data on a millisecond time frame (such as changes in the heart rate of the interviewer and interviewee relative to one another) as each person responds to and interacts with their dyadic partner.

Performer: Charles Stark Draper Laboratory, Inc.

The framework will be based on epidemiological methods used in understanding how infectious diseases are spread throughout a population. The goal is to develop proof-of-concept software tools that can identify the characteristics of “infectious” concepts of extremism/radicalization and how they morph through Internet blogs. The intent is to model how infectious ideas spread and how they evolve over time. Support vector machines will be trained to recognize concepts through the use of words and expressions and then analyze additional textual information (blogs) to track how the concept is replicated.

Performer: SRI International

The research could lay the groundwork for a larger program to automate the management of machine learning algorithms, resulting in dramatic reductions in the need for human experts when working with large, diverse data sets.

Performer: Dartmouth College

Performer: Gamelan Labs, Inc.

Resulting Research: 2016-16032900007

While the IC invests heavily in training advanced language skills, research has demonstrated that language proficiency in adult language learners often does not achieve advanced levels of proficiency and even with the same training proficiency will vary greatly across different adult students. Recent research suggests that general cognitive abilities, brain structure and function may predict language learning outcomes; however, this has never been tested in healthy adults. The Sensors for Predicting Enhanced Language Learning seedling will seek to build upon previous findings to determine if there are cognitive and/or neurophysiological predictors of individuals who are more likely to be successful in learning and retaining a tonal-based novel language. The Massachusetts Institute of Technology research team will assess if there are pre-training features that predict the level of language acquisition throughout the language training, directly following completion, and at a 3 month follow-up.

Performer: Massachusetts Institute of Technology

The idea is to develop a system whereby a cooperative subject would be placed in a non-invasive device (e.g., fMRI) and simply think about an image. The system would then reconstruct the image, identify appropriate attributes, and then search large image databases for similar images. The deliverables would be reports, analyses, and research data (state-of-the-art, program formulation, experimental design, maybe some experimental results). Intelligence analysts would benefit from simply thinking of an image and having a system retrieve similar images rather than entering in image attributes. There would be a significant reduction in image search and retrieval times.

Performer: Johns Hopkins University Applied Physics Laboratory

If successful, the methodology is likely to power new solutions that revolutionize the productivity of the analytical enterprise and the precision and completeness of analytical products.

Performer: University of Southern California

Resulting Research: FA8650-15-C-9102

Recent advances in biotechnology have enabled the rapid, facile engineering of a diversity of biological systems. While these technologies have great promise, they also increase the risk of either accidental or purposeful misuse, and there are currently few methods to detect many sophisticated modifications. This seedling seeks to develop approaches for detecting signatures of biological engineering where the modification method, transgene sequence, and genomic target site(s) are unknown, including current methods of altering gene expression without changing DNA sequence.

Performer: Leidos, Inc.

Resulting Research: 2016-16071200002

As a demonstrator of the techniques, the focus will be on the coding of ground truth data (hereinafter, referred to as the Gold Standard Report) for eight MENA (Middle East/North Africa) countries for four event types: Civil Unrest (CU + Widespread CU events) and Military Action and Non-State Actor (MANSA) events. These events will be coded from news articles gathered from 25 gold standard sources that report articles in English and Arabic.

Performer: Virginia Polytechnic Institute and State University (Virginia Tech)

Resulting Research: N/A

In addition, evaluate the algorithms’ effectiveness and provide program-level insight into 1) a data collection methodology; 2) a methodology and software infrastructure for evaluation; and 3) baseline performance.

Performer: Raytheon BBN Technologies

Resulting Research: N/A

Complex events of the sort of interest to the IC are made up of multiple component events and contextual influences which over time can converge into an action or event of interest. By identifying component events and tracking them over time, event structures can emerge, supporting indicators and warnings to analysts as an input to their task prioritization process.

Performer: Language Computer Corporation

"Model Drift", for this seedling, refers to the discrepancy in machine learning environments between the training (source) and test (target) datasets. Such discrepancies may arise from the gradual or abrupt shift in the characteristics of the input data, or the gradual or abrup invalidation of core assumptions used in the development of the model.

Performer: University of Southern California

Resulting Research: FA8750-15-C-0071

The inputs to this computational pipeline are three correlated data streams collected from autonomous mosquito traps: metagenomic data, wingbeat spectra, and environmental factors. The pipeline is tunable, according to the biological processes that degrade host and pathogen nucleic acids in a time-dependent manner, based on vector conditions. Data from pathogen surveillance programs can lead to better epidemiological models and give health organizations more time to optimally prepare responses.

Performer: Vanderbilt University

Resulting Research: D15PC00304

Complex events of the sort of interest to the IC are made up of multiple component events and contextual influences which over time can converge into an action or event of interest. By identifying component events and tracking them over time, event structures can emerge, supporting indicators and warnings to analysts as an input to their task prioritization process.

Performer: Lymba Corporation

These errors may be unintentional (e.g., malfunctioning sensors) or intentional (e.g., maliciously tampered with, or deleted or obscured for privacy preservation). RIM4S represents multiple error types using a probabilistic graphical model, and then uses belief propagation to efficiently evaluate candidate hypotheses. Current methods to compensate for these types of errors are to threshold and reject data or sensors believed to be inaccurate or unreliable. The proposed approach promises to be able to use degraded information, enabling inference with fewer sensors or measurements.

Performer: University of California, Los Angeles

If successful, the research could lay the groundwork for a larger program to generate predictive models from a wide range of document types (e.g., news articles, speeches, financial reports).

Performer: Carnegie Mellon University

However, multidimensional spectroscopy requires control and measurement of the phases of the electromagnetic waveforms, which while relatively straightforward to accomplish for RF is much more difficult in the optical domain. Demonstrations to date of optical multidimensional spectroscopy require extremely precise optical delay lines and are not suitable for field deployment. Separately, dual comb spectroscopy has developed as a new approach to Fourier Transform optical spectroscopy, without the need for a scanning mechanical delay line. CBMCS seeks to demonstrate multidimensional coherent spectroscopy using three optical frequency combs. The approach under development in the seedling, if successful, could pave the way for development of fieldable devices for chemical detection with high clutter rejection, covert optical communications, and improved synthetic aperture Laser Detection and Ranging (LADAR) waveforms.

Performer: University of Michigan

Resulting Research: 2016-16041300005

Dual-comb spectroscopy is an extremely powerful technique for acquiring high-sensitivity, broadband, and extremely high-resolution optical spectra for chemical identification and quantification. However, despite its sensitivity and resolution advantages, dual-comb spectroscopy is challenging for field implementation, since it requires the repetition rate and phase coherence between two independent laser comb sources to be synchronized and stabilized. This typically requires complex electronic feedback loops and additional reference lasers. In the proposed approach, the two pulse trains constituting the two combs are generated from a common cavity within a passively mode-locked fiber laser. Therefore phase noise from relative intensity noise of the pump laser and the frequency variation due to environmental effects will be predominantly common-mode, enabling phase synchronization between the two optical combs without a complicated and expensive feedback system.

Performer: University of Kansas

Resulting Research: 2017-17030700017

Current non-Foster matching networks use negative capacitors and negative inductors, for which it is a significant challenge to maintain linearity and stability. Using negative resistors should provide fewer parasitic effects, lower current and voltage magnitudes, and improve linearity. The project includes analysis, simulation, design, building, and testing of broadband matching networks for receive and transmit antenna matching. The testing will compare the performance without matching networks and with active matching circuits.

Performer: Northrop Grumman Corporation

Detection of small sequences provides the advantages that a) small segments of pathogenic DNA inserted into otherwise benign sequences are more likely to be detected; b) sequence variations are less likely to lead to missed detection; and c) artificially engineered sequences of concern will be detected by the similarity to small natural segments of pathogenic DNA required to achieve the pathogenic functionality.

Performer: Raytheon BBN Technologies

The long-term goal of the project is to develop a rapid, highly integrated micro-GC system with extremely low power consumption, in which voltage, rather than temperature, is used to control the vapor separation.

Performer: Regents of the University of Michigan

The proposal consists of three main components: 1) synthesis and fabrication of three nonlinear optical (NLO) small organic molecules, 2) grafting those NLO molecules onto fabricated silicon microcavities, and 3) characterization of the grafted NLO microcavity’s performance as a low-threshold microlaser. The central advance of this proposal is the grafting technique by which nonlinear optical small organic molecules can be grafted onto a silicon nano-structure.

Performer: University of Southern California

Resulting Research: 2016-16070100002

To realize the phased array, Northrop will 3D integrate the phase-change switch TDUs with an RF feed-board and a linear array of wide-band 2D notch antennas, resulting in a demonstration of 8-40GHz ±60° directivity.

Performer: Northrop Grumman Systems Corporation

Resulting Research: N/A

1) genes associated with craniofacial development, and 2) CpG lock methylation patterns associated with age estimation.

Performer: Bode Technology Group, Inc.

Resulting Research: D15PC00002

The recent promise of commercial satellites such as WorldView, Planet, and Black Sky Global, with frequent flyovers of the same area (e.g., as many as 72 revisits/day in a few years) and multimodal data capture (e.g., panchromatic, multiple infrared bands), can provide the Intelligence Community with new opportunities to exploit imagery globally.

Performer: SRI International

Resulting Research: N/A

A frequency comb is a coherent light source whose spectrum consists of a series of discrete, equally spaced elements. The SMFC seedling is investigating air-clad, etchless silicon microresonators for comb generation in the MIR. It is also developing fabrication techniques to reduce coupling and propagation losses and optimizing the microresonator dispersion to enable generation of ultra-broadband combs. Potential applications enabled by the SMFC technology include high-resolution spectroscopy, high-speed detection and ultrashort pulse generation.

Performer: Cornell University

Curved focal planes also have the potential to increase image sharpness and increase edge illumination compared to flat field optics. Advancements in the ability to create highly curved infra-red (IR) imaging sensors can enable compact, high-performance, wide field-of-view (FOV) optical systems for low SWaP intelligence, surveillance, and reconnaissance (ISR) applications. HRL has previously demonstrated a unique free- edge compressive bending technique with lithographically thinned commercial off the shelf (COTS) silicon complementary metal oxide semiconductor (CMOS) imaging sensors to create working devices with spherical surfaces exceeding 25° of subtended angle, which is 3X greater than the current state-of-the-art for curved visible-region imaging sensors. The SCIS seedling effort seeks to extend both the compressive bending and patterned strain relief approaches to group III-V Short Wave Infrared (SWIR) and Mid Wave Infrared (MWIR) sensors (including InGaAs, InAs/GaSb and InAsSb).

Performer: Hughes Research Laboratories, Llc.

Resulting Research: 2017-17033000004

Performer: Mitre Corporation

The first stage of the device is a radial sampler with two concentric absorbent bed regions. The “upstream” absorbent bed region is a lower surface area bed which has high capacity and high desorption efficiency for semi-volatiles, while the “downstream” absorbent bed is a higher surface bed, which is better suited for trapping and desorbing more volatile compounds. The combination of these sorbent beds provides efficient absorption across a wide spectrum of sample types and concentrations. After sample collection is complete, thermal desorption liberates the collected analyte from the first stage of the device in 2-5 seconds and “dumps” this desorbed gas into the second stage of the device. In the second stage, progressive heating “rolls up” the vapor sample into a tight bunch, which is then released into the separator. The two stage device is projected to provide injection of the concentrated gas aliquot in ~250 ms, with a conservative projected power requirement per cycle of only 65 J. The device also provides a capability for split injections to further compress the temporal pulse of the emitted gas, and the ability to vent water vapor prior to injection. Room temperature ionic liquids (RTILs) will be tested as an alternative sorbent material, providing the potential for a unique new class of sorbents.

Performer: University of Michigan

Resulting Research: 2017-17012600004

Extending active photonic integrated circuit technology to the ultraviolet and “blue” visible range would enable a much wider range of applications, including ion-trap quantum computing, quantum photonics, visible light communications/interconnects, imaging for forgery analysis, on-chip photoionization, and extremely compact, low power UV Raman spectroscopy (via chip based contact spectroscopy or phased array stand-off sensing). The UV-PC seedling effort will demonstrate the key fundamental components of a UV-visible photonic integrated circuit platform, including waveguides, resonators, grating couplers, electro-optic modulators, phase shifters, Mach-Zehnder switches, and splitters.

Performer: Raytheon BBN Technologies

Resulting Research: 2017-17063000002

We will create a custom hardware design that accelerates the common and expensive operations and integrate that with a lightweight, bare-metal runtime that is purpose-built for graph processing. The prototype system will be effective on dedicated clusters and will form the basis for small, efficient graph accelerators in the future.

Performer: Trustees of Indiana University

Resulting Research: N/A

Where such processes are reliable and unaffected by physical effects such as electromagnetic and mechanical forces and temperature, the change in bulk properties can be utilized as a reliable measure of time or as a chip attribution and identification scheme. Used with integrated circuits, such materials have potential applications for: autonomous, self-timed fuse, chip attribution and identification, and low-power time measurement. This seedling identified and studied suitable transmutational materials, developed CMOS-compatible processes for use in integrated circuits, and identified, characterized, and demonstrated potential applications.

Performer: NanoMason

Resulting Research: N66001-C-13-2004

The goal of this seedling is to develop the general architecture of the RQL Bloom filter circuit family and the cell-level designs of these circuits. The energy efficiency of these designs will be calculated by collecting (through simulation) energy consumption data for Signature Insertion (write) and Query (read &check) operations for different lengths of data streams and different numbers of hash functions per stream.

Performer: Stony Brook University

Resulting Research: 2017-16113000003

For this seedling, software simulations were used to quantify the benefits of a general purpose computer architecture (referred to as "Chakra" - the name for the overall design of this High Performance Computing (HPC) system) to study the reference memory behavior.

Performer: APIC Corporation

Resulting Research: D15PC00202

CCA borrows technology elements from several non-von Neumann computer designs including early work on cellular automata, dataflow, systolic arrays, and more recent work on Asynchronous Multi-Task computing, as well as the PI’s work on the ParalleX execution model and HPX runtime software. The proposal reflects an analysis that suggests that an exascale computer can be developed with today’s CMOS (Complementary Metal-Oxide-Semiconductor) processor technology and managed through dynamic adaptive software techniques.

Performer: Trustees of Indiana University

Resulting Research: N/A

ECR aims to allow for a less-privileged guest instance to monitor the behavior of the hypervisor to detect changes or suspicious behavior.  ECR consists of three core tasks: (1) test framework development (2) development and evaluation of a side-channel sensor and (3) final documentation.

Performer: Assured Information Security

Resulting Research: 2017-16120700002

These so-called dot-product-engines are estimated to enhance speed and power efficiency by orders of magnitude, for many algorithms important to the IC. These algorithms rely on matrix and vector operations, and include but are not limited to pattern recognition in image and video analysis, Fast Fourier Transform, Monte Carlo sampling and optimization, polynomial convolution, streaming data classification, and matrix rank calculation. The goal of this IARPA seedling is to produce specialized, high-power performance analog and digital hardware for performing real vector-matrix multiplication. As is usually the case with analog devices, precision is a key challenge. This effort does not intend to produce devices that perform high-precision computation. Hence the utility of proposed hardware is limited to algorithms where exactness of the computation is not the primary concern.

Performer: Hewlett Packard

In principle, a photon-photon gate offers the ability to switch a data bit at high operating speeds with minimal expended energy. The University of California, San Diego research team is seeking to create an optical parametric mixer that exhibits near-isotropic photon gate response. This mixer will be constructed by measuring the fluctuations along the length of an optical fiber with sub-nanometer accuracy and using this information to synthesize a waveguide with near-ideal response. The seedling project culminates with the demonstration of an all-optical gate that is capable of single-bit switching with fewer than 20 photons at a rate of 100Gbps.

Performer: University of California, San Diego

Photonic PUFs benefit from the greater complexity of optical interactions but, unlike traditional optical PUFs, are realized in planar, chip-scale devices fabricated using standard CMOS processes in a photonic foundry environment. This will result in a highly secure, robust, and low SWaP chip-based hardware fingerprint as well as sources of device-specific, private key material for security applications. For example, Johns Hopkins University envisions these devices as tamper-evident unique identifiers to protect the global defense microelectronic supply chain, as a hardware root-of-trust for critical information processing systems, as unique signatures for tagging and tracking of target systems and collection endpoints across the global digital infrastructure, and as sources of key material for the encryption of information stored on or communicated with mobile devices.

Performer: Johns Hopkins University

Through specific use cases, RAMPARTS will assess apsects of the practical ease of programming, automation of compiling, and effective optimization of FHE computations.

Performer: Galois, Inc.

Resulting Research: 2016-16070700001

The research included using SFQ devices for classical computing and using quantum effects to switch signals.

Performer: University of Southern California

Resulting Research: FA8750-15-C-0203

Efforts in the first year served to identify limitations of single, linear hologram templates working with serial streaming data. This led to the proposal of angular multiplexing to massively scale the technology to search for millions of sub-patterns of interest on >100 GHz data. By redesigning the S2 system to employ spatial + angular multiplexing, together with novel template superposition, it is anticipated the number of programmable sub-patterns can be increased upwards of 1000-fold with 10-100x improvement in power usage per physical channel. By month 15, S2 demonstrated multiplexing of streaming data across 3 angles, with substantial reduction of noise floor and related false positive and negative rates.

Performer: S2 Corporation

VLG1 effort focused on a proof-of-concept demonstration implementing time-dependent graph theory and methods for analyzing very large scale graphs that include temporal dependencies. VLG2 extended the research by studying the impact of uncertainty mechanisms in data on anomalous subgraph detection and to understand and characterize uncertainty in real datasets. VLG-H investigated the advanced hardware system components for processing massive network data two to four orders of magnitude beyond existing capabilities. Specific technical improvements to be explored are data dependent optimization, a high bandwidth communication network, cache-less memory, specialized processing for sparse matrices, low-power circuitry, a sparse array instruction set, and semantic processing.

Performer: Massachusetts Institute of Technology, Lincoln Laboratory