Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “SPEECH”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 37 records · Page 2
All sourcesNASA NTRSDOE OSTI
NASA NTRS1988

A dissociation of objective and subjective workload measures in assessing the impact of speech controls in advanced helicopters

Among the new technologies that are expected to aid helicopter designers are speech controls. Proponents suggest that speech controls could reduce the potential for manual control overloads and improve time-sharing performance in environments that have heavy demands for manual control. This was tested in a simulation of an advanced single-pilot, scout/attack helicopter. Objective performance indicated that the speech controls were effective in decreasing the interference of discrete responses during moments of heavy flight control activity. However, subjective ratings indicated that the use of speech controls required extra effort to speak precisely and to attend to feedback. Although the operational reliability of speech controls must be improved, the present results indicate that reliable speech controls could enhance the time-sharing efficiency of helicopter pilots. Furthermore, the results demonstrated the importance of using multiple assessment techniques to completely assess a task. Neither the objective nor the subjective measures alone provided complete information. It was the contrast between the measures that was most informative.

Vidulich, Michael A.↗
DOE OSTI2022

Speaker-targeted Synthetic Speech Detection

Text-to-speech technologies are evolving quickly towards realistic-sounding human-like voices. As this technology improves, so does the opportunity for malpractice in speaker identification (SID) via spoofing, the process of impersonating a voice biometric via synthesis. More data typically equates to a more realistic voice model, which poses an issue for well-known subjects, such as politicians and celebrities, who have vast amounts of multimedia available online. Detection of synthetic speech has relied on signal processing techniques that focus on the generation of new acoustic features and train deep learning models to detect when an audio file has been manipulated through the characterization of unnatural changes or artifacts. However, these techniques do not use any information from the speaker they are evaluating. This paper proposes to incorporate information from the speaker-of-interest (SoI) into the models to avoid specific spoofing attacks for certain vulnerable people. The wealth of data for well-known people can also be used to train a speaker-specific spoofing detector with a higher level of accuracy than a speaker-independent model. The paper proposes a new xResNet-PLDA system and compares it to three different baseline systems: a state-of-the-art speaker identification system, an xResNet system trained to discriminate between bona fide and fake speech, and a speaker identification system in which the PLDA and calibration models were trained with bona fide and fake speech. We evaluated the systems in two different scenarios — a cross-validation scenario and a hold-out scenario — with three different databases. We show how the proposed system outperforms dramatically the baseline systems in each scenario and for each database. Finally, we show how using a small amount of the SoI’s speech to adapt global calibration parameters improves the performance of the system, especially in unseen conditions.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS2014

Signal Processing Methods for Removing the Effects of Whole Body Vibration upon Speech

Humans may be exposed to whole-body vibration in environments where clear speech communications are crucial, particularly during the launch phases of space flight and in high-performance aircraft. Prior research has shown that high levels of vibration cause a decrease in speech intelligibility. However, the effects of whole-body vibration upon speech are not well understood, and no attempt has been made to restore speech distorted by whole-body vibration. In this paper, a model for speech under whole-body vibration is proposed and a method to remove its effect is described. The method described reduces the perceptual effects of vibration, yields higher ASR accuracy scores, and may significantly improve intelligibility. Possible applications include incorporation within communication systems to improve radio-communication systems in environments such a spaceflight, aviation, or off-road vehicle operations.

human response to vibration↗
DOE OSTI2019

Deep learning as a tool for neural data analysis: Speech classification and cross-frequency coupling in human sensorimotor cortex

A fundamental challenge in neuroscience is to understand what structure in the world is represented in spatially distributed patterns of neural activity from multiple single-trial measurements. This is often accomplished by learning a simple, linear transformations between neural features and features of the sensory stimuli or motor task. While successful in some early sensory processing areas, linear mappings are unlikely to be ideal tools for elucidating nonlinear, hierarchical representations of higher-order brain areas during complex tasks, such as the production of speech by humans. Here, we apply deep networks to predict produced speech syllables from a dataset of high gamma cortical surface electric potentials recorded from human sensorimotor cortex. We find that deep networks had higher decoding prediction accuracy compared to baseline models. Having established that deep networks extract more task relevant information from neural data sets relative to linear models (i.e., higher predictive accuracy), we next sought to demonstrate their utility as a data analysis tool for neuroscience. We first show that deep network's confusions revealed hierarchical latent structure in the neural data, which recapitulated the underlying articulatory nature of speech motor control. We next broadened the frequency features beyond high-gamma and identified a novel high-gamma-to-beta coupling during speech production. Finally, we used deep networks to compare task-relevant information in different neural frequency bands, and found that the high-gamma band contains the vast majority of information relevant for the speech prediction task, with little-to-no additional contribution from lower-frequency amplitudes. Together, these results demonstrate the utility of deep networks as a data analysis tool for basic and applied neuroscience.

60 APPLIED LIFE SCIENCES↗
NASA NTRS1972

Asynchronous sampling of speech with some vocoder experimental results

The method of asynchronously sampling speech is based upon the derivatives of the acoustical speech signal. The following results are apparent from experiments to date: (1) It is possible to represent speech by a string of pulses of uniform amplitude, where the only information contained in the string is the spacing of the pulses in time; (2) the string of pulses may be produced in a simple analog manner; (3) the first derivative of the original speech waveform is the most important for the encoding process; (4) the resulting pulse train can be utilized to control an acoustical signal production system to regenerate the intelligence of the original speech.

Babcock, M. L.↗
NASA NTRS1977

Effects of interior aircraft noise on speech intelligibility and annoyance

Recordings of the aircraft ambiance from ten different types of aircraft were used in conjunction with four distinct speech interference tests as stimuli to determine the effects of interior aircraft background levels and speech intelligibility on perceived annoyance in 36 subjects. Both speech intelligibility and background level significantly affected judged annoyance. However, the interaction between the two variables showed that above an 85 db background level the speech intelligibility results had a minimal effect on annoyance ratings. Below this level, people rated the background as less annoying if there was adequate speech intelligibility.

Pearsons, K. S.↗
NASA NTRS1982

Measurement of speech levels in the presence of time varying background noise

Short-term speech level measurements which could be used to note changes in vocal effort in a time varying noise environment were studied. Knowing the changes in speech level would in turn allow prediction of intelligibility in the presence of aircraft flyover noise. Tests indicated that it is possible to use two second samples of speech to estimate long term root mean square speech levels. Other tests were also performed in which people read out loud during aircraft flyover noise. Results of these tests indicate that people do indeed raise their voice during flyovers at a rate of about 3-1/2 dB for each 10 dB increase in background level. This finding is in agreement with other tests of speech levels in the presence of steady state background noise.

Pearsons, K. S.↗
NASA NTRS1984

Speech communications in noise

The physical characteristics of speech, the methods of speech masking measurement, and the effects of noise on speech communication are investigated. Topics include the speech signal and intelligibility, the effects of noise on intelligibility, the articulation index, and various devices for evaluating speech systems.

Source record↗
NASA NTRS1986

Versatile simulation testbed for rotorcraft speech I/O system design

A versatile simulation testbed for the design of a rotorcraft speech I/O system is described in detail. The testbed will be used to evaluate alternative implementations of synthesized speech displays and speech recognition controls for the next generation of Army helicopters including the LHX. The message delivery logic is discussed as well as the message structure, the speech recognizer command structure and features, feedback from the recognizer, and random access to controls via speech command.

Simpson, Carol A.↗
NASA NTRS1989

Pulse Vector-Excitation Speech Encoder

Proposed pulse vector-excitation speech encoder (PVXC) encodes analog speech signals into digital representation for transmission or storage at rates below 5 kilobits per second. Produces high quality of reconstructed speech, but with less computation than required by comparable speech-encoding systems. Has some characteristics of multipulse linear predictive coding (MPLPC) and of code-excited linear prediction (CELP). System uses mathematical model of vocal tract in conjunction with set of excitation vectors and perceptually-based error criterion to synthesize natural-sounding speech.

Davidson, Grant↗
NASA NTRS1990

Comparison of speech intelligibility in cockpit noise using SPH-4 flight helmet with and without active noise reduction

Active Noise Reduction (ANR) is a new technology which can reduce the level of aircraft cockpit noise that reaches the pilot's ear while simultaneously improving the signal to noise ratio for voice communications and other information bearing sound signals in the cockpit. A miniature, ear-cup mounted ANR system was tested to determine whether speech intelligibility is better for helicopter pilots using ANR compared to a control condition of ANR turned off. Two signal to noise ratios (S/N), representative of actual cockpit conditions, were used for the ratio of the speech to cockpit noise sound pressure levels. Speech intelligibility was significantly better with ANR compared to no ANR for both S/N conditions. Variability of speech intelligibility among pilots was also significantly less with ANR. When the stock helmet was used with ANR turned off, the average PB Word speech intelligibility score was below the Normally Acceptable level. In comparison, it was above that level with ANR on in both S/N levels.

Chan, Jeffrey W.↗
NASA NTRS1990

A variable rate speech compressor for mobile applications

One of the most promising speech coder at the bit rate of 9.6 to 4.8 kbits/s is CELP. Code Excited Linear Prediction (CELP) has been dominating 9.6 to 4.8 kbits/s region during the past 3 to 4 years. Its set back however, is its expensive implementation. As an alternative to CELP, the Base-Band CELP (CELP-BB) was developed which produced good quality speech comparable to CELP and a single chip implementable complexity as reported previously. Its robustness was also improved to tolerate errors up to 1.0 pct. and maintain intelligibility up to 5.0 pct. and more. Although, CELP-BB produces good quality speech at around 4.8 kbits/s, it has a fundamental problem when updating the pitch filter memory. A sub-optimal solution is proposed for this problem. Below 4.8 kbits/s, however, CELP-BB suffers from noticeable quantization noise as a result of the large vector dimensions used. Efficient representation of speech below 4.8 kbits/s is reported by introducing Sinusoidal Transform Coding (STC) to represent the LPC excitation which is called Sine Wave Excited LPC (SWELP). In this case, natural sounding good quality synthetic speech is obtained at around 2.4 kbits/s.

Yeldener, S.↗
NASA NTRS1991

The effects of speech controls on performance in advanced helicopters in a double stimulation paradigm

The potential benefit of speech as a control modality has been investigated with mixed results. Earlier studies suggests that speech controls can reduce the potential of manual control overloads and improve time-sharing performance. However, these benefits were not without costs. Pilots reported higher workload levels associated with the use of speech controls. To further investigate these previous findings, an experiment was conducted in a simulation of an advanced single-pilot, scout/attack helicopter at NASA-Ames' ICAB (interchangeable cab) facility. Objective performance data suggested that speech control modality was effective in reducing interference of discrete, time-shared responses during continuous flight control activity. Subjective ratings, however, indicated that the speech control modality increased workload. Post-flight debriefing indicated that these results were mainly due to the increased effort to speak precisely to a less than perfect voice recognition system.

Bortolussi, Michael R.↗
NASA NTRS2015

Automatic Speech Acquisition and Recognition for Spacesuit Audio Systems

NASA has a widely recognized but unmet need for novel human-machine interface technologies that can facilitate communication during astronaut extravehicular activities (EVAs), when loud noises and strong reverberations inside spacesuits make communication challenging. WeVoice, Inc., has developed a multichannel signal-processing method for speech acquisition in noisy and reverberant environments that enables automatic speech recognition (ASR) technology inside spacesuits. The technology reduces noise by exploiting differences between the statistical nature of signals (i.e., speech) and noise that exists in the spatial and temporal domains. As a result, ASR accuracy can be improved to the level at which crewmembers will find the speech interface useful. System components and features include beam forming/multichannel noise reduction, single-channel noise reduction, speech feature extraction, feature transformation and normalization, feature compression, and ASR decoding. Arithmetic complexity models were developed and will help designers of real-time ASR systems select proper tasks when confronted with constraints in computational resources. In Phase I of the project, WeVoice validated the technology. The company further refined the technology in Phase II and developed a prototype for testing and use by suited astronauts.

Ye, Sherry↗
NASA NTRS2016

Memo on Speech Alarms: Replication and Validation of Results

Caution and warning (C&W) alarms help people to quickly and efficiently identify situations that are of immediate danger or would escalate to a safety critical level. Tones are highly salient and have been traditionally used for caution and warning alarms. However, research shows that tone alarms can have an unwanted startle effect that hinders operator decision making. Speech alarms are good alternatives to tone alarms because they require less training and are less startling. They have been in use for decades for caution and warning systems in commercial airplanes and in buildings. Speech alarms have been considered for space flight use by the National Aeronautics and Space Administration’s (NASA) Astronaut Office and by its Orion Program. To investigate whether performance with various types of speech alarms was similar to performance with the currently used tone alarms, a study was conducted in 2010. The results showed faster identification times of speech alarms as well as higher acceptance rates from participants. However, the presentation of the alarms had a variable onset time due to software. The current research project was funded to address this issue by collecting new data with alarms having nonvariable onset time and to validate the alarms in the Human Exploration Research Analog (HERA). This report describes the two studies: a laboratory experiment comparing tone and speech alarms, and an evaluation in the HERA facility.

warning signals↗
NASA NTRSDate not supplied

Transcribing Air Traffic Control System Command Center Planning Telecons Using Cloud-Based Automatic Speech Recognition

This paper addresses the challenge of using Automatic Speech Recognition (ASR) technology to transcribe regular teleconferences that happen between FAA Air Traffic Control System Command Center (ATCSCC) planners, stakeholders and air users. These planning teleconferences (aka telecons or planning webinars) are an integral part of managing air traffic in the U.S. National Airspace System (NAS). In particular, the meetings facilitate the creation and modification of various traffic management initiatives (TMIs), that are used to regulate the flow of air traffic. This is typically a human intensive process, requiring specialists to listen to the entire meeting audio (10-20 minutes duration) and inferring the state of the NAS (e.g., weather phenomenon) that was discussed. It would be advantageous to have digital transcripts of the audio and have useful information (e.g., related to TMIs) automatically extracted from the transcripts. In this regard, we are exploring the adoption of state-of-the-art speech to text and Natural Language Processing (NLP) tools that will achieve our objective of digitizing the webinar audio. Unfortunately, the highly technical phraseology present in the audio and limited data availability for model building make ASR difficult. To overcome this challenge, we have taken the critical first step in creating a human transcription dataset from ~20 hours of speech in the ATCSCC audio with the help of subject matter experts. A novelty of our work is the creation of a ground truth transcription dataset for ATCSCC teleconference webinars, which is particularly important for Aviation domain-specific NLP tasks. Using Microsoft Speech Studio, a cloud-based ASR platform, we have fine-tuned the English pre-trained ASR models (available in speech studio) and achieved an average word error rate (WER) of 6.81%. The baseline ASR also provides a digital version of each planning webinar, making it accessible and text-searchable for future references. Additionally, the transcriptions can serve as a bridge between raw audio data and a range of text-based NLP tasks, such as named entity recognition (NER) and intent classification, potentially enhancing the digital footprint of the webinars and other connected data sources. Our work has several potential applications. Firstly, the transcriptions can be analyzed to understand the complex decision process of creating, implementing and modifying TMIs and may also contribute to TMI prediction services. Secondly, our dataset and model can be used to develop more accurate ASR systems for aviation-specific language, which can bring about digital communication in the aviation industry (and aid current “voice only” communications, which are inherently error-prone). Lastly, the transcriptions themselves can be used as a valuable resource for training other NLP models.

Stephen S. B. Clarke↗
NASA NTRS2015

Speech Intelligibility Advantages using an Acoustic Beamformer Display

A speech intelligibility test conforming to the Modified Rhyme Test of ANSI S3.2 "Method for Measuring the Intelligibility of Speech Over Communication Systems" was conducted using a prototype 12-channel acoustic beamformer system. The target speech material (signal) was identified against speech babble (noise), with calculated signal-noise ratios of 0, 5 and 10 dB. The signal was delivered at a fixed beam orientation of 135 deg (re 90 deg as the frontal direction of the array) and the noise at 135 deg (co-located) and 0 deg (separated). A significant improvement in intelligibility from 57% to 73% was found for spatial separation for the same signal-noise ratio (0 dB). Significant effects for improved intelligibility due to spatial separation were also found for higher signal-noise ratios (5 and 10 dB).

acoustic beamforming↗
NASA NTRS2002

Speech motor control and acute mountain sickness

BACKGROUND: An objective method that accurately quantifies the severity of Acute Mountain Sickness (AMS) symptoms is needed to enable more reliable evaluation of altitude acclimatization and testing of potentially beneficial interventions. HYPOTHESIS: Changes in human articulation, as quantified by timed variations in acoustic waveforms of specific spoken words (voice onset time; VOT), are correlated with the severity of AMS. METHODS: Fifteen volunteers were exposed to a simulated altitude of 4300 m (446 mm Hg) in a hypobaric chamber for 48 h. Speech motor control was determined from digitally recorded and analyzed timing patterns of 30 different monosyllabic words characterized as voiced and unvoiced, and as labial, alveolar, or velar. The Environmental Symptoms Questionnaire (ESQ) was used to assess AMS. RESULTS: Significant AMS symptoms occurred after 4 h, peaked at 16 h, and returned toward baseline after 48 h. Labial VOTs were shorter after 4 and 39 h of exposure; velar VOTs were altered only after 4 h; and there were no changes in alveolar VOTs. The duration of vowel sounds was increased after 4 h of exposure and returned to normal thereafter. Only 1 of 15 subjects did not increase vowel time after 4 h of exposure. The 39-h labial (p = 0.009) and velar (p = 0.037) voiced-unvoiced timed separations consonants and the symptoms of AMS were significantly correlated. CONCLUSIONS: Two objective measures of speech production were affected by exposure to 4300 m altitude and correlated with AMS severity. Alterations in speech production may represent an objective measure of AMS and central vulnerability to hypoxia.

NASA Discipline Space Human Factors↗