Publications
An overview of automatic speaker recognition technology
May 13, 2002
Conference Paper
Published in:
Proc. IEEE Int. Conf. on Acoustics, Speech and Signal Processing, ICASSP, Vol. IV, 13-17 May 2002, pp. IV-4072 - IV-4075.
Topic:
R&D area:
Summary
In this paper we provide a brief overview of the area of speaker recognition, describing applications, underlying techniques and some indications, of performance. Following this overview we will discuss some of the strengths and weaknesses of current speaker recognition technologies and outline some potential future trends in research, development and applications conducting other speech interactions (background verification). As speaker and speech recognition system merge and speech recognition accuracy improves, the distinction between text- independent and -dependent applications will decrease. Of the two basic tasks, text-dependent speaker verification is currently
Summary
In this paper we provide a brief overview of the area of speaker recognition, describing applications, underlying techniques and some indications, of performance. Following this overview we will discuss some of the strengths and weaknesses of current speaker recognition technologies and outline some potential future trends in research, development and...
READ MORE
Speaker verification using text-constrained Gaussian mixture models
May 13, 2002
Conference Paper
Published in:
Proc. IEEE Int. Conf. on Acoustics, Speech and Signal Processing, ICASSP, Vol. I, 13-17 May 2002, pp. I-677 - I-680.
Topic:
R&D area:
Summary
In this paper we present an approach to close the gap between text-dependent and text-independent speaker verification performance. Text-constrained GMM-UBM systems are created using word segmentations produced by a LVCSR system on conversational speech allowing the system to focus on speaker differences over a constrained set of acoustic units. Results on the 2001 NiST extended data task show this approach can be used to produce an equal error rate of < 1%.
Summary
In this paper we present an approach to close the gap between text-dependent and text-independent speaker verification performance. Text-constrained GMM-UBM systems are created using word segmentations produced by a LVCSR system on conversational speech allowing the system to focus on speaker differences over a constrained set of acoustic units. Results...
READ MORE
Speaker detection and tracking for telephone transactions
May 13, 2002
Conference Paper
Published in:
Proc. IEEE Int. Conf. on Acoustics, Speech, and Signal Processing, ICASSP, 13-17 May 2002, pp. 129-132.
Topic:
R&D area:
Summary
As ever greater numbers of telephone transactions are being conducted solely between a caller and an automated answering system, the need increases for software which can automatically identify and authenticate these callers without the need for an onerous speaker enrollment process. In this paper we introduce and investigate a novel speaker detection and tracking (SDT) technique, which dynamically merges the traditional enrollment and recognition phases of the static speaker recognition task. In this speaker recognition application, no prior speaker models exist and the goal is to detect and model new speakers as they call into the system while also recognizing utterances from the previously modeled callers. New speakers are added to the enrolled set of speakers and speech from speakers in the currently enrolled set is used to update models. We describe a system based on a GMM speaker identification (SID) system and develop a new measure to evaluate the performance of the system on the SDT task. Results for both static, open-set detection and the SDT task are presented using a portion of the Switchboard corpus of telephone speech communications. Static open-set detection produces an equal error rate of about 5%. As expected, performance for SDT is quite varied, depending greatly on the speaker set and ordering of the test sequence. These initial results, however, are quite promising and point to potential areas in which to improve the system performance.
Summary
As ever greater numbers of telephone transactions are being conducted solely between a caller and an automated answering system, the need increases for software which can automatically identify and authenticate these callers without the need for an onerous speaker enrollment process. In this paper we introduce and investigate a novel...
READ MORE
Speech enhancement based on auditory spectral change
May 13, 2002
Conference Paper
Published in:
Proc. IEEE Int. Conf. on Acoustics, Speech and Signal Processing, ICASSP, Vol. I, Speech Processing Neural Networks for Signal Processing, 13-17 May 2002, pp. I-257 - I-260.
Topic:
R&D area:
Summary
In this paper, an adaptive approach to the enhancement of speech signals is developed based on auditory spectral change. The algorithm is motivated by sensitivity of aural biologic systems to signal dynamics, by evidence that noise is aurally masked by rapid changes in a signal, and by analogies to these two aural phenomena in biologic visual processing. Emphasis is on preserving nonstationarity, i.e., speech transient and time-varying components, such as plosive bursts, formant transitions, and vowel onsets, while suppressing additive noise. The essence of the enhancement technique is a Wiener filter that uses a desired signal spectrum whose estimation adapts to stationarity of the measured signal. The degree of stationarity is derived from a signal change measurement, based on an auditory spectrum that accentuates change in spectral bands. The adaptive filter is applied in an unconventional overlap-add analysis/synthesis framework, using a very short 4-ms analysis window and a 1-ms frame interval. In informal listening, the reconstructions are judged to be "crisp" corresponding to good temporal resolution of transient and rapidly-moving speech events.
Summary
In this paper, an adaptive approach to the enhancement of speech signals is developed based on auditory spectral change. The algorithm is motivated by sensitivity of aural biologic systems to signal dynamics, by evidence that noise is aurally masked by rapid changes in a signal, and by analogies to these...
READ MORE
Automated generation and analysis of attack graphs
May 12, 2002
Conference Paper
Published in:
Proc. of the 2002 IEEE Symp. on Security and Privacy, 12-15 May 2002, pp. 254-265.
Topic:
R&D area:
R&D group:
Summary
An integral part of modeling the global view of network security is constructing attack graphs. In practice, attack graphs are produced manually by Red Teams. Construction by hand, however, is tedious, error-prone, and impractical for attack graphs have larger than a hundred nodes. In this paper we present an automated technique for generating and analyzing attack graphs. We base our technique on symbolic model checking algorithms, letting us construct attack graphs automatically and efficiently. We also describe two analyses to help decide which attacks would be most cost-effective to guard against. We implemented our techniques in a tool suite and tested it on a small network example, which includes models of a firewall and an intrusion detection system.
Summary
An integral part of modeling the global view of network security is constructing attack graphs. In practice, attack graphs are produced manually by Red Teams. Construction by hand, however, is tedious, error-prone, and impractical for attack graphs have larger than a hundred nodes. In this paper we present an automated...
READ MORE
Speech-to-speech translation: technology and applications study
May 10, 2002
Technical Report
Published in:
MIT Lincoln Laboratory Report TR-1080
Topic:
R&D area:
Summary
This report describes a study effort on the state-of-the-art and lessons learned in automated, two- way, speech-to-speech translation and its potential application to military problems. The study includes and comments upon an extensive set of references on prior and current work in speech translation. The study includes recommendations on future military applications and on R&D needed to successfully achieve those applications. Key findings of the study include: (1) R&D speech translation systems have been demonstrated, but only in limited domains, and their performance is inadequate for operational use; (2) as far as we have been able to determine, there are currently no operational two-way speech translation systems; (3) intensive, sustained R&D will be needed to develop usable two-way speech translation systems. Major recommendations include: (1) a substantial R&D program in speech translation is needed, especially including full end-to-end system prototyping and evaluation; (2) close cooperation among researchers and users speaking multiple languages will be needed for the development of useful application systems; (3) to get military users involved and interacting in a mode which enables them to provide useful inputs and feedback on system requirements and performance, it will be necessary to provide them at the start with a fairly robust, open-domain system which works to the degree that some two-way speech translation is operational.
Summary
This report describes a study effort on the state-of-the-art and lessons learned in automated, two- way, speech-to-speech translation and its potential application to military problems. The study includes and comments upon an extensive set of references on prior and current work in speech translation. The study includes recommendations on future...
READ MORE
Time domain processing of frequency domain GPR signatures for buried land mine detection
April 1, 2002
Conference Paper
Published in:
SPIE, Vol. 4742, Part One, Detection and Remediation Technologies for Mines and Minelike Targets VII, 1-5 April 2002, pp. 339-348.
Summary
This paper investigates the feasibility of detecting plastic antipersonnel land mines buried in lossy, dispersive, rough soils using a stepped-frequency ultra wideband (WB) ground-penetrating radar (GPR). Realistic land mine scenarios were modeled using a two-dimensional (2D) finite difference firequency domain (FDFD) technique. Assuming normal incidence plane wave excitation, the scattered fields were generated over a large frequency bandwidth (.5 to 5 GHz) for a variety of mine-like shapes, different soil types, and multiple receiver locations. The simulation results showed that for a ground penetration sensor located just above the soil surface, the strong reflection signals received from the rough ground surface obscured the buried target's fiquency response signal. The simulated GPR WB frequency response data at each receiver location was transformed to the time domain using the fast fourier transform. Time domain processing permits high resolution measurement of target features that are invariant to the ground roughness and also that are dependent on the soil characteristics as well as the burial depth and size of the mine, Specifically, two or more characteristic timing peaks are observed in the simulation results suggesting that the ultra-wideband spectral radar response may yield particular advantages not exploited by currently employed detection systems. It is also shown that by using time-gating to remove the strong ground reflection signals, the target signals are selectively enhanced (as expected), but more surprisingly, the target frequency response signature is almost completely recovered.
Summary
This paper investigates the feasibility of detecting plastic antipersonnel land mines buried in lossy, dispersive, rough soils using a stepped-frequency ultra wideband (WB) ground-penetrating radar (GPR). Realistic land mine scenarios were modeled using a two-dimensional (2D) finite difference firequency domain (FDFD) technique. Assuming normal incidence plane wave excitation, the scattered...
READ MORE
COTS fusion tracker evaluation
February 15, 2002
Project Report
Published in:
MIT Lincoln Laboratory Report ATC-302
Summary
Lincoln Laboratory was tasked by the FAA to measure the performance of a representative sample of current commercial off-the-shelf (COTS) fusion trackers. This effort included cataloging the companies that have available ATC fusion trackers, acquiring executable tracker images from as many as possible of these trackers, running the commercial tracker code on the test sets, and evaluating the performance achieved. This report presents an overall review of the state-of-the-art of fusion tracker as applied to the FAA surveillance problem. Average statistics of performance, as well as performance in special situations, are included. In each case, the performance of fusion is compared against the performance of single sensor and mosaic tracking. Thus, the advantages and disadvantages of fusion will be evident. The statistics may also permit the generation of a fusion tracker specification should the FAA decide to procure one as part of a future automation system.
Summary
Lincoln Laboratory was tasked by the FAA to measure the performance of a representative sample of current commercial off-the-shelf (COTS) fusion trackers. This effort included cataloging the companies that have available ATC fusion trackers, acquiring executable tracker images from as many as possible of these trackers, running the commercial tracker...
READ MORE
New products for the NEXRAD ORPG to support FAA critical systems
February 9, 2002
Conference Paper
Published in:
19th Int. Conf. on Interactive Processing Systems for Meteorology, Oceanography and Hydrology, 9-13 February 2002.
Topic:
R&D area:
R&D group:
Summary
A number of Federal Aviation Administration (FAA) critical systems rely on products from the NEXRAD (WSR-88D) suite of algorithms. These systems include MIAWS (Medium Intensity Airport Weather System), ITWS (Integrated Terminal Weather System), CIWS (Corridor Integrated Weather System), and WARP (Weather and Radar Processing). With the advent of the NEXRAD Open Radar Product Generator (ORPG), a six-month build cycle has been established for the incorporation of new or improved algorithms. This build cycle provides the mechanism for the integration of new products into the algorithm suite tailored to the needs of these FAA systems now and into the future. Figure 1 is useful for visualizing the MIT/LL ORPGnet. Four of the ORPGnet systems are located at MIT/LL headquartered in Lexington, MA. These four systems form the core of the development center where algorithms are developed for and implemented into the ORPG environment. Part of the development process includes examination of algorithm products created from past weather. A number of utilities are available for playback of various versions of NEXRAD Archive II base data: from tape or disk files in standard or LDM formats. Additionally, MIT/LL operates the CIWS demonstation project for the FAA. The ORPG clones at the development center have access to base data from 26 NEXRAD radars from the Midwest to the East Coast of the United States ingested for CIWS. The FAA has tasked the Massachusetts Institute of Technology's Lincoln Laboratory (MIT/LL) with developing algorithms for the ORPG to address their systems' needs. Many of these algorithms will also prove useful to other users of NEXRAD products such as the National Weather Service and the Department of Defense. MIT/LL has created a network of ten ORPGs, or an ORPGnet, to use for the purpose of developing, testing, and implementing new algorithms targeted to specific builds. The benefits of the ORPGnet will be discussed in more detail later in this paper. MIT/LL has provided improvements to existing algorithms or developed new algorithms for the first three build cycles of the ORPG (Istok et al., 2002; Smalley and Bennett, 2002). Development of more algorithms is currently in progress for upcoming build cycles. In addition to describing ORPGnet, this paper will focus on its use in the development of a new Data Quality Assurance (DQA) algorithm, an improved High Resolution VIL (HRVIL) algorithm, and progress on the development of the enhanced Echo Tops (EET) algorithm; as well as the symbiotic relationship of these algorithms to the FAA critical systems.
Summary
A number of Federal Aviation Administration (FAA) critical systems rely on products from the NEXRAD (WSR-88D) suite of algorithms. These systems include MIAWS (Medium Intensity Airport Weather System), ITWS (Integrated Terminal Weather System), CIWS (Corridor Integrated Weather System), and WARP (Weather and Radar Processing). With the advent of the NEXRAD...
READ MORE
Contributions to the AIAA Guidance, Navigation & Control Conference
January 23, 2002
Project Report
Published in:
MIT Lincoln Laboratory Report NASA-A-5
Topic:
R&D area:
R&D group:
Summary
This report contains six papers presented by the Lincoln Laboratory Air Traffic Control Systems Group at the American Institute of Aeronautics & Astronautics (AIAA) Guidance, Navigation and Control (GNC) conference on 6-9 August 2001 in Montreal, Canada. The work reported was sponsored by the NASA Advanced Air Transportation Technologies (AATT) program and the FAA Free Flight Phase 1 (FFPl) program. The papers are based on studies completed at Lincoln Laboratory in collaboration with staff at NASA Ames Research Center. These papers were presented in the Air Traffic Automation Session of the conference and fall into three major areas: Traffic Analysis & Benefits Studies, Weather/Automation Integration, and Surface Surveillance. In the first area, a paper by Andrews & Robinson presents an analysis of the efficiency of runway operations at Dallas/l%. Worth using a tool called PARO, and a paper by Welch, Andrews, & Robinson presents delay benefit results for the Final Approach Spacing Tool (FAST). In the second area, a paper by Campbell, et al. describes a new weather distribution system for the Center/TRACON Automation System (CTAS) that allows ingestion of multiple weather sources, and a paper by van de Venne, Lloyd, & Hogaboom describes the use of the NOAA Eta model as a backup wind data source for CTAS. Also in this area, a paper by Murphy & Campbell presents initial steps towards integrating weather-impacted routes into FAST. In the third area, a paper by Welch, Bussolari, and Atkins presents an initial operational concept for using surface surveillance to reduce taxi delays.
Summary
This report contains six papers presented by the Lincoln Laboratory Air Traffic Control Systems Group at the American Institute of Aeronautics & Astronautics (AIAA) Guidance, Navigation and Control (GNC) conference on 6-9 August 2001 in Montreal, Canada. The work reported was sponsored by the NASA Advanced Air Transportation Technologies (AATT)...
READ MORE