Publications

Refine Results

(Filters Applied) Clear All

Speech recognition by humans and machines under conditions with severe channel variability and noise

Published in:
SPIE, Vol. 3077, Applications and Science of Artificial Neural Networks III, 21-24 April 1997, pp. 46-57.

Summary

Despite dramatic recent advances in speech recognition technology, speech recognizers still perform much worse than humans. The difference in performance between humans and machines is most dramatic when variable amounts and types of filtering and noise are present during testing. For example, humans readily understand speech that is low-pass filtered below 3 kHz or high-pass filtered above 1kHz. Machines trained with wide-band speech, however, degrade dramatically under these conditions. An approach to compensate for variable unknown sharp filtering and noise is presented which uses mel-filter-bank magnitudes as input features, estimates the signal-to-noise ratio (SNR) for each filter, and uses missing feature theory to dynamically modify the probability computations performed using Gaussian Mixture or Radial Basis Function neural network classifiers embedded within Hidden Markov Model (HMM) recognizers. The approach was successfully demonstrated using a talker-independent digit recognition task. It was found that recognition accuracy across many conditions rises from below 50 % to above 95 % with this approach. These promising results suggest future work to dynamically estimate SNR's and to explore the dynamics of human adaptation to channel and noise variability.
READ LESS

Summary

Despite dramatic recent advances in speech recognition technology, speech recognizers still perform much worse than humans. The difference in performance between humans and machines is most dramatic when variable amounts and types of filtering and noise are present during testing. For example, humans readily understand speech that is low-pass filtered...

READ MORE

Analysis of downstream impacts of air traffic delay

Published in:
MIT Lincoln Laboratory Report ATC-257

Summary

Reduction of air carrier flight delay in the U.S. National Airspace System (NAS) has been a major objective of the Federal Aviation Administration (FAA) for many years. Much of the current delay arises from weather-induced delays at airports. When a plane is delayed on one of the day's flights, there can be a carryover delay that affects later flights by that aircraft. In this report, we develop statistical models to predict: 1. The "downstream" delays that occur when a flight experiences an initial delay, and 2. The likelihood of flight cancellation as a function of the initial delay. Using historical airline-reported delays for December 1993, we conclude that the mean "downstream" delay is approximately 80 percent of the initial delay, i.e., the net delay for an aircraft due to an initial flight delay is approximately 1.8 x the initial delay.
READ LESS

Summary

Reduction of air carrier flight delay in the U.S. National Airspace System (NAS) has been a major objective of the Federal Aviation Administration (FAA) for many years. Much of the current delay arises from weather-induced delays at airports. When a plane is delayed on one of the day's flights, there...

READ MORE

Convective weather forecasting for FAA applications

Published in:
7th Conf. on Aviation, Range, and Aerospace Meteorology, ARAM, 2-7 February 1997.

Summary

The Convective Weather Product Development Team (PDT) was formed in 1996 as part of the reorganization of the FAA Aviation Weather Research Program, to provide an effective way to conduct critical applied research in a collaborative and rational fashion. Detecting and predicting convective weather is extremely important to aviation, since approximately half of the national airspace delay in the warm season is caused by thunderstorms. Reliable 0--6 hr storm predictions are essential for aviation users to achieve safe and efficient use of the airspace, as well as for future air traffic control automation systems. Our goal on this PDT is to direct our research and development activities toward operationally useful convective weather detection and forecast products, and delivery of those products, so that users can receive benefits on an immediate and continual basis. Given that we have many more initiatives than funding, we have chosen to prioritize our activities according to near-term achievable benefits to users. Our hope is that the success of initial planned demonstrations will help the FAA identify a consistent level of long-term R&D funding, so that we can make real progress towards achieving our full set of goals. In this paper, we present our statement of the FAA Convective Weather Forecasting problem, evidence of the need for forecasts in the National Airspace System (NAS), and an illustration of the air traffic delay caused by convective weather. We then discuss our research plan and rationale, and outline our main initiatives for the upcoming year.
READ LESS

Summary

The Convective Weather Product Development Team (PDT) was formed in 1996 as part of the reorganization of the FAA Aviation Weather Research Program, to provide an effective way to conduct critical applied research in a collaborative and rational fashion. Detecting and predicting convective weather is extremely important to aviation, since...

READ MORE

The impact of thunderstorm growth and decay on air traffic management in class B airspace

Published in:
7th Conf. on Aviation, Range, and Aerospace Meteorology, ARAM, 2-7 February 1997.

Summary

Air traffic management is a challenging task, especially if the airspace involved is impacted by inclement weather. The high volume of air traffic which inundates the nation's major airports compounds the difficulties with which Air Traffic Control (ATC) specialists have to cope. When you add the unpredictability of thunderstorm growth and decay to the controllers workload, air traffic management becomes even more of a challenge. ATC specialists would benefit from reliable forecasts of thunderstorm growth and decay. To determine how they would use a Growth and Decay product, ATC specialists from the Memphis Air Route Traffic Control Center (ARTCC), Traffic Management Unit (TMU), and TRACON supervisors were interviewed while viewing five movie loops of Memphis weather cases. The movies consisted of the ASR-9 six-level reflectivity data, aircraft beacons, and storm motion vectors.
READ LESS

Summary

Air traffic management is a challenging task, especially if the airspace involved is impacted by inclement weather. The high volume of air traffic which inundates the nation's major airports compounds the difficulties with which Air Traffic Control (ATC) specialists have to cope. When you add the unpredictability of thunderstorm growth...

READ MORE

The Memphis ITWS convective forecasting collaborative demonstration

Summary

Accurate, short-term forecasts of where thunderstorms will develop, move and decay allow for strategic traffic management in and around the aviation terminal and enroute airspace. Pre-planning to avoid adverse weather conditions provides safe, smooth and continuous air traffic flow and savings in both fuel cost and time. Wolfson, et. al ( 1997) describe the problem of convective weather forecasting for FAA applications. In 1995, National Center for Atmospheric Research (NCAR), MIT Lincoln Laboratory (MIT-LL) and National Severe Storms Laboratory (NSSL) scientists and engineers agreed to collaborate on the development of a convective weather forecasting algorithm for use in airport terminal areas. Each laboratory brings special strengths to the project. NCAR has been developing techniques for precise, short-term (0-60 minutes) forecasts of thunderstorm initiation, movement and dissipation for the FAA over the past ten years and has developed the Auto-Nowcaster software. MIT-LL has been developing real-time algorithms for the Integrated Terminal Weather System (ITWS), including techniques for storm tracking, gust front detection, and calculating storm growth and decay (as part of predicting microbursts) . NSSL has been working on the NEXRAD Storm Cell Identification and Tracking (SCIT) algorithm, and on understanding the predictive value of the storm cell information. Thus by using the latest research results and best techniques available at each laboratory, the collaborative effort will hopefully result in a superior convective weather forecasting algorithm. Our goal in the immediate future is to develop a joint algorithm that can be demonstrated to users of terminal weather information, so that the benefits of convective weather forecast information can be realized, and the remaining needs can be assessed. As a first effort in the collaboration, the laboratories fielded their individual algorithms at the Memphis ITWS site. This paper gives an overview of our collaborative experiment in Memphis, the system each laboratory operated, some preliminary analysis of our performance on one case, and our plans for the near future.
READ LESS

Summary

Accurate, short-term forecasts of where thunderstorms will develop, move and decay allow for strategic traffic management in and around the aviation terminal and enroute airspace. Pre-planning to avoid adverse weather conditions provides safe, smooth and continuous air traffic flow and savings in both fuel cost and time. Wolfson, et. al...

READ MORE

Beacon radar and TCAS reply rates: airborne measurements in the 1090 MHz band

Published in:
MIT Lincoln Laboratory Report ATC-256

Summary

The Federal Aviation Administration (FAA) is in the process of developing Automatic Dependent Surveillance Broadcast (ADS-B) techniques. In one candidate system, GPS-Squitter, each aircraft periodically broadcasts messages, called "squitters," that report the aircraft's identification, position, and velocity. The position and velocity information may be obtained from the Global Positioning System (GPS) or some other navigation device. Reception of squitters can be used for several purposes, including surveillance of airborne aircraft by a ground station, surveillance of aircraft on the airport surface, and air-to-air surveillance... In developing the new system, it is necessary to know the rates of existing signal transmissions in the 1030 and 1090 MHz frequency bands, which are the beacon-radar and TCAS interrogation channels. The GPS-Squitter would be transmitted in the 1090 MHz band, like a reply. A key issue is the possibility of interference to squitter reception from existing signals in the 1090 MHz band....To validate these initial calculations, Lincoln Laboratory has made direct measurements of the rates of existing transmissions in both bands. These signals consist mainly of interrogations in the 1030 MHz band and replies in the 1090 MHz band. This report focuses on airborne measurements that have been made at 1090 MHz. (Not complete)
READ LESS

Summary

The Federal Aviation Administration (FAA) is in the process of developing Automatic Dependent Surveillance Broadcast (ADS-B) techniques. In one candidate system, GPS-Squitter, each aircraft periodically broadcasts messages, called "squitters," that report the aircraft's identification, position, and velocity. The position and velocity information may be obtained from the Global Positioning System...

READ MORE

Report on product performance for the Terminal Doppler Weather Radars (TDWRs) at Washington National Airport and Memphis and Orlando International Airports

Published in:
MIT Lincoln Laboratory Report ATC-246

Summary

Massachusetts Institute of Technology Lincoln Laboratory provides support to the Terminal Doppler Weather Radar (TDWR) Program Office in the performance analysis of deployed TDWR systems, and resulting recommendations for systems enhancements. This report documents initial performance of the TDWR products at Washington National Airport (DCA), Memphis International Airport (MEM) and Orlando International Airport (MCO). This performance depends, in turn, on the site optimization performed for the specific radars. Therefore, an overview of site optimization process, using DCA as a concrete example, is included. After the sites were optimized, base data (Doppler velocity and reflectivity) and product data (algorithm detections) were collected to assess the quality of the base data and the performance of the microburst and gust front detection algorithms. It is assumed that the reader of this report has an extensive knowledge of the TDWR system. (Not Complete)
READ LESS

Summary

Massachusetts Institute of Technology Lincoln Laboratory provides support to the Terminal Doppler Weather Radar (TDWR) Program Office in the performance analysis of deployed TDWR systems, and resulting recommendations for systems enhancements. This report documents initial performance of the TDWR products at Washington National Airport (DCA), Memphis International Airport (MEM) and...

READ MORE

Terminal area separation standards: historical development, current standards, and processes for change

Published in:
MIT Lincoln Laboratory Report ATC-258

Summary

This paper gives an overview and summary of the separation requirements for air traffic control in the U.S. National Airspace System with emphasis on those relevant to terminal landing operations. These requirements are documented in the Federal Aviation Administration's (FAA's) Air Traffic Control Order 7110.65J, as ammended, and various national and local Orders. These requirements are also addressed in the Aeronautical Information Manual, the International Civil Aviation Organization's Standards and Recommended Practices, and the Federal Aviation Regulations (FARs). The purpose of this paper is to assist those people involved with the introduction of new technologies and procedures in the terminal airspace by providing them with an understanding of the separation requirements, the need for those requirements, and the processes used to change the requirements.
READ LESS

Summary

This paper gives an overview and summary of the separation requirements for air traffic control in the U.S. National Airspace System with emphasis on those relevant to terminal landing operations. These requirements are documented in the Federal Aviation Administration's (FAA's) Air Traffic Control Order 7110.65J, as ammended, and various national...

READ MORE

AM-FM separation using auditory-motivated filters

Published in:
IEEE Trans. Speech Audio Process., Vol. 5, No. 5, September 1997, pp. 465-480.

Summary

An approach to the joint estimation of sine-wave amplitude modulation (AM) and frequency modulation (FM) is described based on the transduction of frequency modulation into amplitude modulation by linear filters, being motivated by the hypothesis that the auditory system uses a similar transduction mechanism in measuring sine-wave FM. An AM-FM estimation is described that uses the amplitude envelope of the output of two transduction filters of piecewise-linear spectral shape. The piecewise-linear constraint is then relaxed, allowing a wider class of transduction-filter pairs for AM-FM separation under a monotonicity constraint of the filters' quotient. The particular case of Gaussian filters, and measured auditory filters, although not leading to a solution in closed form, provide for iterative AM-FM estimation. Solution stability analysis and error evaluation are performed and the FM transduction method is compared with the energy separation algorithm, based on the Teager energy operator, and the Hilbert transform method for AM-FM estimation. Finally, a generalization to two-dimensional (2-D) filters is described.
READ LESS

Summary

An approach to the joint estimation of sine-wave amplitude modulation (AM) and frequency modulation (FM) is described based on the transduction of frequency modulation into amplitude modulation by linear filters, being motivated by the hypothesis that the auditory system uses a similar transduction mechanism in measuring sine-wave FM. An AM-FM...

READ MORE

Sub-Poisson statistics observed in an electronically shuttered and back-illuminated CCD pixel

Author:
Published in:
IEEE Trans. Electron Devices, Vol. 44, No. 1, January 1997, pp. 69-73. Selected Papers on CCD and CMOS Imagers, SPIE Milestone Series, Vol. MS 177, 2003, pp. 169-173.

Summary

The variance versus average signal has been measured for a pixel in an electronically shuttered and back-illuminated CCD imaging array. The measurements demonstrate that, over a certain operating range, the electronic shutter modifies the input Poisson distributed photoelectrons during the collection process such that the charge signal accumulated in the CCD well has a sub-Poisson distribution (variance less than a mean). A simple one-dimensional model has been developed that explains the experimental results.
READ LESS

Summary

The variance versus average signal has been measured for a pixel in an electronically shuttered and back-illuminated CCD imaging array. The measurements demonstrate that, over a certain operating range, the electronic shutter modifies the input Poisson distributed photoelectrons during the collection process such that the charge signal accumulated in the...

READ MORE