Publications

Refine Results

(Filters Applied) Clear All

Benchmarking the MIT LL HPCMP DHPI system

June 19, 2007
  |
Conference Paper
Author:
Albert I. Reuther
…
Published in:
Annual High Performance Computer Modernization Program Users Group Conf., 19-21 June 2007.
Topic:
high performance computing
R&D area:
R&D group:

Summary

The Massachusetts Institute of Technology Lincoln Laboratory (MIT LL) High Performance Computing Modernization Program (HPCMP) Dedicated High Performance Computing Project Investment (DHPI) system was designed to address interactive algorithm development for Department of Defense (DoD) sensor processing systems. The results of the system acceptance test provide a clear quantitative picture of the capabilities of the system. The system acceptance test for MIT LL HPCMP DHPI hardware involved an array of benchmarks that exercised each of the components of the memory hierarchy, the scheduler, and the disk arrays. These benchmarks isolated the components to verify the functionality and performance of the system, and several system issues were discovered and rectified by using these benchmarks. The memory hierarchy was evaluated using the HPC Challenge benchmark suite, which is comprised of the following benchmarks: High Performance Linpack (HPL, also known as Top 500), Fast Fourier Transform (FFT), STREAM, RandomAccess, and Effective Bandwidth. The compute nodes' Random Array of Independent Disks (RAID) arrays were evaluated with the Iozone benchmark. Finally, the scheduler and the reliability of the entire system were tested using both the HPC Challenge suite and the Iozone benchmark. For example executing the HPC Challenge benchmark suite on 416 processors, the system was able to achieve 1.42 TFlops (HPL), 34.7 GFlops (FFT), 1.24 TBytes/sec (STREAM Triad), and 0.16 GUPS (RandomAccess). This paper describes the components of the MIT Lincoln Laboratory HPCMP DHPI system, including its memory hierarchy. We present the HPC Challenge benchmark suite and Iozone benchmark and describe how each of the component benchmarks stress various components of the TX-2500 system. The results of the benchmarks are discussed, and the implications they have on the performance of the system. We conclude with a presentation of the findings.
READ LESS

Summary

The Massachusetts Institute of Technology Lincoln Laboratory (MIT LL) High Performance Computing Modernization Program (HPCMP) Dedicated High Performance Computing Project Investment (DHPI) system was designed to address interactive algorithm development for Department of Defense (DoD) sensor processing systems. The results of the system acceptance test provide a clear quantitative picture...

READ MORE
Benchmarking the MIT LL HPCMP DHPI system

Automatic language identification

June 15, 2007
  |
Journal Article
Author:
Pedro A. Torres-Carrasquillo
…
Published in:
Wiley Encyclopedia of Electrical and Electronics Engineering, Vol. 2, pp. 104-9, 2007.
Topic:
language recognition
R&D area:
R&D group:

Summary

Automatic language identification is the process by which the language of digitized spoken words is recognized by a computer. It is one of several processes in which information is extracted automatically from a speech signal.
READ LESS

Summary

Automatic language identification is the process by which the language of digitized spoken words is recognized by a computer. It is one of several processes in which information is extracted automatically from a speech signal.

READ MORE
Automatic language identification

Making network intrusion detection work with IPsec

May 11, 2007
  |
Technical Report
Author:
Cynthia D. McLain
…
Published in:
MIT Lincoln Laboratory Report TR-1121
Topic:
cryptography
R&D area:
R&D group:

Summary

Network-based intrusion detection systems (NIDSs) are one component of a comprehensive network security solution. The use of IPsec, which encrypts network traffic, renders network intrusion detection virtually useless unless traffic is decrypted at network gateways. One alternative to NIDSs, host-based intrusion detection systems (HIDSs), provides some of the functionality of NIDSs but with limitations. HIDSs cannot perform a network-wide analysis and can be subverted if a host is compromised. We propose an approach to intrusion detection that combines HIDS, NIDS, and a version of IPsec that encrypts the header and the body of IP packets separately. We refer to the latter generically as Two-Key IPsec. We show that all of the network events currently detectable by the Snort NIDS on unencrypted network traffic are also detectable on encrypted network traffic using this approach. The NIDS detects network-level events that HIDSs have trouble detecting and HIDSs detect application-level events that can't be detected by the NIDS.
READ LESS

Summary

Network-based intrusion detection systems (NIDSs) are one component of a comprehensive network security solution. The use of IPsec, which encrypts network traffic, renders network intrusion detection virtually useless unless traffic is decrypted at network gateways. One alternative to NIDSs, host-based intrusion detection systems (HIDSs), provides some of the functionality of...

READ MORE
Making network intrusion detection work with IPsec

MIT Lincoln Laboratory multimodal person identification system in the CLEAR 2007 Evaluation

May 8, 2007
  |
Conference Paper
Author:
Kevin Brady
Published in:
2nd Annual Classification of Event Activities and Relationships/Rich Transcription Evaluations, 8-11 May 2008, pp. 240-247.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

A description of the MIT Lincoln Laboratory system used in the person identification task of the recent CLEAR 2007 Evaluation is documented in this paper. This task is broken into audio, visual, and multimodal subtasks. The audio identification system utilizes both a GMM and a SVM subsystem, while the visual (face) identification system utilizes an appearance-based [Kernel] approach for identification. The audio channels, originating from a microphone array, were preprocessed with beamforming and noise preprocessing.
READ LESS

Summary

A description of the MIT Lincoln Laboratory system used in the person identification task of the recent CLEAR 2007 Evaluation is documented in this paper. This task is broken into audio, visual, and multimodal subtasks. The audio identification system utilizes both a GMM and a SVM subsystem, while the visual...

READ MORE
MIT Lincoln Laboratory multimodal person identification system in the CLEAR 2007 Evaluation

Low-bit-rate speech coding

May 1, 2007
  |
Book Chapter
Author:
Alan V. McCree
Published in:
Chapter 16 in Springer Handbook of Speech Processing and Communication, 2007, pp. 331-50.
Topic:
speech processing
R&D area:
R&D group:

Summary

Low-bit-rate speech coding, at rates below 4 kb/s, is needed for both communication and voice storage applications. At such low rates, full encoding of the speech waveform is not possible; therefore, low-rate coders rely instead on parametric models to represent only the most perceptually relevant aspects of speech. While there are a number of different approaches for this modeling, all can be related to the basic linear model of speech production, where an excitation signal drives a vocal-tract filter. The basic properties of the speech signal and of human speech perception can explain the principles of parametric speech coding as applied in early vocoders. Current speech modeling approaches, such as mixed excitation linear prediction, sinusoidal coding, and waveform interpolation, use more-sophisticated versions of these same concepts. Modern techniques for encoding the model parameters, in particular using the theory of vector quantization, allow the encoding of the model information with very few bits per speech frame. Successful standardization of low-rate coders has enabled their widespread use for both military and satellite communications, at rates from 4 kb/s all the way down to 600 b/s. However, the goal of toll-quality low-rate coding continues to provide a research challenge.
READ LESS

Summary

Low-bit-rate speech coding, at rates below 4 kb/s, is needed for both communication and voice storage applications. At such low rates, full encoding of the speech waveform is not possible; therefore, low-rate coders rely instead on parametric models to represent only the most perceptually relevant aspects of speech. While there...

READ MORE
Low-bit-rate speech coding

Nuisance attribute projection

May 1, 2007
  |
Book Chapter
Author:
Alex Solomonoff
…
Published in:
Chapter in Speech Communication, May 2007.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

Cross-channel degradation is one of the significant challenges facing speaker recognition systems. We study this problem in the support vector machine (SVM) context and nuisance variable compensation in high-dimensional spaces more generally. We present an approach to nuisance variable compensation by removing nuisance attribute-related dimensions in the SVM expansion space via projections. Training to remove these dimensions is accomplished via an eigenvalue problem. The eigenvalue problem attempts to reduce multisession variation for the same speaker, reduce different channel effects, and increase "distance" between different speakers. Experiments show significant improvement in performance for the cross-channel case.
READ LESS

Summary

Cross-channel degradation is one of the significant challenges facing speaker recognition systems. We study this problem in the support vector machine (SVM) context and nuisance variable compensation in high-dimensional spaces more generally. We present an approach to nuisance variable compensation by removing nuisance attribute-related dimensions in the SVM expansion space...

READ MORE
Nuisance attribute projection

Text-independent speaker recognition

May 1, 2007
  |
Book Chapter
Author:
Douglas A. Reynolds
…
Published in:
Springer Handbook of Speech Processing and Communication, 2007, pp. 763-81.
Topic:
speaker recognition
R&D area:
R&D group:

Summary

In this chapter, we focus on the area of text-independent speaker verification, with an emphasis on unconstrained telephone conversational speech. We begin by providing a general likelihood ratio detection task framework to describe the various components in modern text-independent speaker verification systems. We next describe the general hierarchy of speaker information conveyed in the speech signal and the issues involved in reliably exploiting these levels of information for practical speaker verification systems. We then describe specific implementations of state-of-the-art text-independent speaker verification systems utilizing low-level spectral information and high-level token sequence information with generative and discriminative modeling techniques. Finally, we provide a performance assessment of these systems using the National Institute of Standards and Technology (NIST) speaker recognition evaluation telephone corpora.
READ LESS

Summary

In this chapter, we focus on the area of text-independent speaker verification, with an emphasis on unconstrained telephone conversational speech. We begin by providing a general likelihood ratio detection task framework to describe the various components in modern text-independent speaker verification systems. We next describe the general hierarchy of speaker...

READ MORE
Text-independent speaker recognition

ILR-based MT comprehension test with multi-level questions

April 22, 2007
  |
Conference Paper
Author:
Douglas A. Jones
…
Published in:
Human Language Technology, North American Chapter of the Association for Computational Linguistics, HLT/NAACL, 22-27 April 2007.
Topic:
machine translation
R&D area:
R&D group:

Summary

We present results from a new Interagency Language Roundtable (ILR) based comprehension test. This new test design presents questions at multiple ILR difficulty levels within each document. We incorporated Arabic machine translation (MT) output from three independent research sites, arbitrarily merging these materials into one MT condition. We contrast the MT condition, for both text and audio data types, with high quality human reference Gold Standard (GS) translations. Overall, subjects achieved 95% comprehension for GS and 74% for MT, across all genres and difficulty levels. Interestingly, comprehension rates do not correlate highly with translation error rates, suggesting that we are measuring an additional dimension of MT quality.
READ LESS

Summary

We present results from a new Interagency Language Roundtable (ILR) based comprehension test. This new test design presents questions at multiple ILR difficulty levels within each document. We incorporated Arabic machine translation (MT) output from three independent research sites, arbitrarily merging these materials into one MT condition. We contrast the...

READ MORE
ILR-based MT comprehension test with multi-level questions

A new approach to achieving high-performance power amplifier linearization

April 17, 2007
  |
Conference Paper
Author:
Joel I. Goodman
…
Published in:
IEEE Radar Conf., 17-20 April 2007. doi: 10.1109/RADAR.2007.374329
Topic:
signal processing
R&D area:
R&D group:

Summary

Digital baseband predistortion (DBP) is not particularly well suited to linearizing wideband power amplifiers (PAs); this is due to the exorbitant price paid in computational complexity. One of the underlying reasons for the computational complexity of DBP is the inherent inefficiency of using a sufficiently deep memory and a high enough polynomial order to span the multidimensional signal space needed to mitigate PA-induced nonlinear distortion. Therefore we have developed a new mathematical method to efficiently search for and localize those regions in the multidimensional signal space that enable us to invert PA nonlinearities with a significant reduction in computational complexity. Using a wideband code division multiple access (CDMA) signal we demonstrate and compare the PA linearization performance and computational complexity of our algorithm to that of conventional DBP techniques using measured results.
READ LESS

Summary

Digital baseband predistortion (DBP) is not particularly well suited to linearizing wideband power amplifiers (PAs); this is due to the exorbitant price paid in computational complexity. One of the underlying reasons for the computational complexity of DBP is the inherent inefficiency of using a sufficiently deep memory and a high...

READ MORE
A new approach to achieving high-performance power amplifier linearization

Language recognition with word lattices and support vector machines

April 15, 2007
  |
Conference Paper
Author:
William M. Campbell
…
Published in:
Proc. IEEE Int. Conf. on Acoustics, Speech, and Signal Processing, ICASSP, 15-20 April 2007, Vol. IV, pp. 989-992.
Topic:
language recognition
R&D area:
R&D group:

Summary

Language recognition is typically performed with methods that exploit phonotactics--a phone recognition language modeling (PRLM) system. A PRLM system converts speech to a lattice of phones and then scores a language model. A standard extension to this scheme is to use multiple parallel phone recognizers (PPRLM). In this paper, we modify this approach in two distinct ways. First, we replace the phone tokenizer by a powerful speech-to-text system. Second, we use a discriminative support vector machine for language modeling. Our goals are twofold. First, we explore the ability of a single speech-to-text system to distinguish multiple languages. Second, we fuse the new system with an SVM PRLM system to see if it complements current approaches. Experiments on the 2005 NIST language recognition corpus show the new word system accomplishes these goals and has significant potential for language recognition.
READ LESS

Summary

Language recognition is typically performed with methods that exploit phonotactics--a phone recognition language modeling (PRLM) system. A PRLM system converts speech to a lattice of phones and then scores a language model. A standard extension to this scheme is to use multiple parallel phone recognizers (PPRLM). In this paper, we...

READ MORE
Language recognition with word lattices and support vector machines
521-530 of 809