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 253 records · Page 14

Intelligent interfaces for expert systems

Vital to the success of an expert system is an interface to the user which performs intelligently. A generic intelligent interface is being developed for expert systems. This intelligent interface was developed around the in-house developed Expert System for the Flight Analysis System (ESFAS). The Flight Analysis System (FAS) is comprised of 84 configuration controlled FORTRAN subroutines that are used in the preflight analysis of the space shuttle. In order to use FAS proficiently, a person must be knowledgeable in the areas of flight mechanics, the procedures involved in deploying a certain payload, and an overall understanding of the FAS. ESFAS, still in its developmental stage, is taking into account much of this knowledge. The generic intelligent interface involves the integration of a speech recognizer and synthesizer, a preparser, and a natural language parser to ESFAS. The speech recognizer being used is capable of recognizing 1000 words of connected speech. The natural language parser is a commercial software package which uses caseframe instantiation in processing the streams of words from the speech recognizer or the keyboard. The systems configuration is described along with capabilities and drawbacks.

Villarreal, James A.↗

A 4800 bps CELP vocoder with an improved excitation

The Stochastic or Code Excited Linear Predictive Coder (CELP) is among the promising candidates for producing good quality speech at low bit rates. However, the speech quality produced suffers from perceived roughness. Many researchers have used pole-zero postfilters to mask the roughness at the output of the synthesis filter. Although the postfilters are effective in masking the noise at low bit rates, they produce spectral distortions. It is proposed that speech can be improved by introducing two modifications to the fixed stochastic codebook. In the first modification, the stochastic codebook is used only when the long term correlations are low. Otherwise, a pulse like codebook is selected. In the second modification, the selected codebook output is weighted using an adaptive spectral shaping procedure. These two modifications were incorporated in a 4800 bps CELP coder and have resulted in a perceptually improved vocoded speech.

Hassanein, Hisham↗

A low-delay 8 Kb/s backward-adaptive CELP coder

Code excited linear prediction coding is an efficient technique for compressing speech sequences. Communications quality of speech can be obtained at bit rates below 8 Kb/s. However, relatively large coding delays are necessary to buffer the input speech in order to perform the LPC analysis. A low delay 8 Kb/s CELP coder is introduced in which the short term predictor is based on past synthesized speech. A new distortion measure that improves the tracking of the formant filter is discussed. Formal listening tests showed that the performance of the backward adaptive coder is almost as good as the conventional CELP coder.

Neumeyer, L. G.↗

Voice integrated systems

The program at Naval Air Development Center was initiated to determine the desirability of interactive voice systems for use in airborne weapon systems crew stations. A voice recognition and synthesis system (VRAS) was developed and incorporated into a human centrifuge. The speech recognition aspect of VRAS was developed using a voice command system (VCS) developed by Scope Electronics. The speech synthesis capability was supplied by a Votrax, VS-5, speech synthesis unit built by Vocal Interface. The effects of simulated flight on automatic speech recognition were determined by repeated trials in the VRAS-equipped centrifuge. The relationship of vibration, G, O2 mask, mission duration, and cockpit temperature and voice quality was determined. The results showed that: (1) voice quality degrades after 0.5 hours with an O2 mask; (2) voice quality degrades under high vibration; and (3) voice quality degrades under high levels of G. The voice quality studies are summarized. These results were obtained with a baseline of 80 percent recognition accuracy with VCS.

Curran, P. Mike↗

Communication indices of crew coordination

The relationship between communication patterns and performance in 10 two-person flightcrews is explored with the aim of identifying speech variations which differentiate low- and high-error full mission simulator flights. Verbal data, transcribed from the videotaped performances, are treated as interactive sequences of speech events in which statements spoken by one crewmember are considered within the context of the other crewmember's prior and subsequent speech. Specific speech patterns characterized each crew, but the overriding findings included: a) marked homogeneity of patterns characterizing low-error crews, interpreted as the adoption of a standard form of communicating, and b) heterogeneity of patterns characterizing high-error crews, interpreted as the relative absence of a conventionalized form. Because conventions are regularities which confirm the expectations of those involved, predictability of crewmember behavior should be greater when standard conventions are followed. We conclude that such a practice can facilitate the coordination process and enhance crew performance.

Aerospace Medicine↗

Enabling a Voice Management System for Space Applications, Design and Software Development

Sustainable missions, beyond low Earth orbit, will require autonomous capabilities in order to achieve NASA’s Artemis program objectives. Correspondingly, the crew must have a means to efficiently interact with these autonomous systems; this can be facilitated via voice and speech communications. Voice-based controls enable the user to access autonomous systems hands-free/eyes-free, allowing the user to better focus on critical tasks. The goal of this project was to explore the knowledge and technology needed to successfully design effective voice interfaces for autonomous systems. The main objective was to understand how a crew member, through voice interaction, could most efficiently and intuitively communicate with a notional autonomous vehicle system manager. This project leveraged prior research conducted by the University of Michigan’s Bioastronautics and Life Support System (BLiSS) team as part of a NASA Moon to Mars eXploration Systems and Habitation (M2M X-Hab) 2020 Academic Innovation Challenge. The X-Hab 2020 work from the BliSS Team resulted in an intuitive graphical user interface/user experience that was built on an Internet of Things (IOT) platform. The Voice User Interface (VUI) design for the M2M X-Hab 2021 project leveraged this technology and incorporated a voice-based assistant and NASA’s Platform for Autonomous Systems (NPAS) software. This required technologies to convert voice to text, conduct semantic interpretations, and convert responses from the autonomous system to text and to speech; additionally, the background noise environment of spacecraft was assessed, and a relatable personality for the autonomous system to facilitate human-like conversations was created. This work’s success was largely due to the diverse team that included expertise in Space Systems Engineering, Human Computer Interaction, Aerospace Engineering, Computer Science, Biomedical Engineering, and Applied Physics. The differing perspectives fostered elaborate discussions, resulting in the conception of three main interactions: (1) User-System, (2) NPAS-System, and (3) Environment-System. The system developed, i.e. the VUI, had to be unique, efficient, and intuitive; thus, the team crafted a personality for the system to enable human-like conversation. User surveys sent to students and young professionals were used to help determine these personality traits by capturing perspectives and expectations of the “Artemis Generation Astronauts”. To further simulate human-like conversations, the system had to be able to quickly interpret user speech and be able to integrate with NASA’s NPAS system for quick and reliable information transfer. Results of this research include (1) a working prototype user interface, that is compatible with NASA’s NPAS system; (2) software that demonstrates the ability to interpret user requests and respond appropriately; (3) the capability to implement fully expanded conversations between user and system using intuitive communication in four request categories; and (4) software and hardware recommendations that optimize the system’s ability to operate, i.e. be heard, in a noisy environment. The technologies chosen for this project’s demonstrations included the following: Raspberry Pi, RASA, Mozilla Deep Speech, Coqui, RTX Voice and Adobe XD. This work has laid the foundation for the development of VUI’s used for autonomy, and is intended to provide guidance for future VUI development.

Tara Vega↗

Enabling a Voice Management System for Space Applications

The sustainable missions beyond Low Earth Orbit (LEO) envisioned for NASA’s Artemis program will require autonomous capabilities. Moreover, Artemis mission crews will need a means to efficiently interact with a spacecraft’s autonomous systems. This interaction can be facilitated by voice and speech communications because voice-based controls enable users to interact hands- and eyes-free, allowing the user to better focus on critical tasks. The goal of our project was to explore the knowledge and technology needed to successfully design effective Voice User Interfaces (VUIs) for autonomous systems utilizing Human Centered Design (HCD) principles. The focus of the human factors’ aspect of engineering, pays close attention to psychological and physiological principles in the development of autonomous crew operation systems. A main objective was to understand how a crew member, through voice interaction, could efficiently and intuitively communicate with a notional autonomous vehicle system manager. This project was a part of the NASA Moon to Mars eXploration Systems and Habitation (M2M X-Hab) 2020 Academic Innovation Challenge. The work from the BLiSS Team, at the University of Michigan, resulted in the design of a system persona, Diego, to which an astronaut may quickly build trust with autonomous systems, to alleviate known stressors on mental health expected during long duration space missions. Optimal software to facilitate integration of the system persona into a reference Lunar orbiting Gateway station was defined. Additionally, a Speech to Text (STT) system and a Graphical User Interface (GUI) that could be implemented in future missions was developed on an Internet of Things (IOT) platform. The Voice User Interface (VUI) design for the M2M X-Hab 2020 project leveraged previous technology developed by the BLiSS team to incorporate a voice-based interface into NASA’s Platform for Autonomous Systems (NPAS) software. This required technologies to convert voice to text, conduct semantic interpretations, and convert responses from the autonomous system to text and to speech; additionally, the spacecraft background noise environment was assessed, a noise mitigation technique was developed, and a relatable personality for the autonomous system was developed in order to facilitate human-like conversations. The success of our effort was largely due to the diversity of the team that included expertise in Space Systems Engineering, Human Computer Interaction, Aerospace Engineering, Computer Science, Biomedical Engineering, and Applied Physics. The diverse perspectives fostered elaborate discussions, resulting in the conception of three main subsystems: (1) User-System, (2) NPAS-System, and (3) Environment-System. The VUI was unique and had to be efficient and intuitive. For this project, 5 subteams were formed, each with a separate objective, Voice Design team, Background Noise Mitigation team, Software Integration team and Graphical User Interface team. The BLiSS team crafted a personality for the VUI to enable human-like conversation and drive user adoption and trust. User surveys were completed and used to help determine the required VUI system personality traits by capturing perspectives and expectations of prospective “Artemis Generation Astronauts”. To further simulate human-like conversations, the system had to be able to quickly interpret user speech and be able to integrate with NASA’s NPAS platform for quick and reliable information transfer. The outcomes of our research were: (1) a working prototype user interface, that is compatible with NASA’s NPAS platform; (2) software that demonstrates the ability of the VUI system to interpret user requests and respond appropriately; (3) the capability to implement fully expanded conversations between user and system using intuitive communication in four request categories; and (4) software and hardware recommendations that optimize the system’s ability to operate in a noisy environment. Our research has laid the foundation for the development of VUI’s for autonomy, and provides a baseline for future VUI developments.

Voice user interface↗

Joint speaker authentication and key phrase identification

A spoken command analyzer computing system includes technologies configured to analyze information extracted from a speech sample and, using a joint speaker and phonetic content model, both determine whether the analyzed speech includes certain content (e.g., a command) and to identify the identity of the human speaker of the speech. In response to determining that the identity matches the authorized user's identity and determining that the analyzed speech includes the modeled content (e.g., command), an action corresponding to the verified content (e.g., command) is performed by an associated device.

97 MATHEMATICS AND COMPUTING↗

A blind algorithm for recovering articulator positions from acoustics

MIMICRI is a signal-processing algorithm that has been shown to blindly infer and invert memoryless nonlinear functions of unobservable bandlimited signals, such as the mapping from the unobservable positions of the speech articulators to observable speech sounds. We review results of using MIMICRI on toy problems and on human speech data. We note that MIMICRI requires that the user specify two parameters: the dimensionality and pass-band of the unobservable signals. We show how to use cross-validation to help estimate the passband. An unexpected consequence of this work is that it helps separate signals with overlapping frequency bands.

42 ENGINEERING↗

Atoms for peace and the nonproliferation treaty: unintended consequences

In April 2009, President Obama revived nonproliferation and arms control efforts with a speech calling for the worldwide abolition of nuclear weapons. His speech correctly acknowledged the threat of nuclear terrorism and the vulnerabilities of the related unsecure nuclear materials. Unfortunately, the president did not mention and has not mentioned in any speech the threat posed by at-risk radiological materials. Nonproliferation efforts have a well documented history of focus on special nuclear materials (fissionable weapons usable materials or SNM), and other key materials (chemical and biological) and technologies for a Weapon of Mass Destruction (WMD). Such intense focus on WMD related materials/technologies is essential for international safety and security and merit continued attention and funding. However, the perception that radioactive sealed sources (sources) are of less concern than WMD is unfortunate. These perceptions are based solely on the potentially enormous and tragic consequences associated with their deliberate or accidental misuse and proliferation concerns. However, there is a documented history of overemphasis on the nuclear threat at the expense of ignoring the far more likely and also devastating chemical and biological threats. The radiological threat should not be minimized or excluded from policy discussions and decisions on these far ranging scopes of threat to the international community. Sources have a long history of use; and a wider distribution worldwide than fissile materials. Pair this with their broad ranges in isotopes/activities along with scant national and international attention and mechanisms for their safe and secure management and it is not difficult to envision a deadly threat. Arguments that minimize or divert attention away from sources may have the effect of distracting necessary policy attention on preventing/mitigating a radiological dispersal event. The terrorist attacks on 9/11 should be a clear reminder of the inherent danger of diminishing or dismissing lower-level threats in exchange for enhanced focus on high priority special nuclear materials with the basis for this emphasis being solely on the magnitude of the consequences of a single event. Mitigating all possible or likely terrorist attacks is impossible; however, weaponized sources, in the form of a radiological dispersal device, have been a declared target material of Al-Qaida. Eisenhower's Atoms for Peace initiative promoted the spread of the paradoxical beneficial yet destructive properties of the atom. Typically, the focus of nonproliferation efforts focuses on the fissile materials associated with Weapons of Mass Destruction, with less emphasis on radioactive materials that could be used for a Weapon of Mass Disruption. Most nonproliferation policy discussion involves securing or preventing the diversion of weapons grade fissile materials (uranium (U) with concentration of over 90% of the isotope {sup 235}U (HEU) and plutonium with more than 90% of the isotope {sup 239}Pu), with scant attention given to the threat posed by a prolific quantity of sources spread worldwide. Further acerbating the problem of inattention, it appears that the momentum of the continued evolution in the beneficial applications of sources will only increase in the near future. Several expert studies have demonstrated on the potentially devastating economic, psychological and public health impacts of terrorist use of a radiological dispersal or radiation emitting device (ROD/RED) in a metropolis. The development of such a weapon, from the acquisition of the radioactive material to the technical knowledge needed to fashion it into an ROD, is many orders of magnitude easier than diverting enough fissile material for and fabrication/acquisition of a nuclear weapon. Unlike nuclear weapons, worldwide, there are many well documented accounts of accidental and purposeful diversions of radioactive materials from regulatory control. As of the end of 2008, the International Atomic Energy Agency's (IAEA) Illicit Trafficking Database had logged 1562 incidents, of which only 18 include weapons grade nuclear materials. As much as 66% of the radioactive material involved in these incidents was not recovered. Since 2004, there has been a 75% increase in incidents of unrecoverable material, much of which is labeled dangerous with potential for deterministic health affects if misused. This makes clear that a black market of illicit trade in sources exists. The incidents reported to the IAEA's database rely only on voluntary state reporting; therefore, the number of lost or stolen sources is expected to be much higher.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

Invariance to background noise as a signature of non-primary auditory cortex

Despite well-established anatomical differences between primary and non-primary auditory cortex, the associated representational transformations have remained elusive. Here we show that primary and non-primary auditory cortex are differentiated by their invariance to real-world background noise. We measured fMRI responses to natural sounds presented in isolation and in real-world noise, quantifying invariance as the correlation between the two responses for individual voxels. Non-primary areas were substantially more noise-invariant than primary areas. This primary-nonprimary difference occurred both for speech and non-speech sounds and was unaffected by a concurrent demanding visual task, suggesting that the observed invariance is not specific to speech processing and is robust to inattention. The difference was most pronounced for real-world background noise—both primary and non-primary areas were relatively robust to simple types of synthetic noise. Our results suggest a general representational transformation between auditory cortical stages, illustrating a representational consequence of hierarchical organization in the auditory system.

59 BASIC BIOLOGICAL SCIENCES↗

Kinship structures create persistent channels for language transmission

Here, languages are transmitted through channels created by kinship systems. Given sufficient time, these kinship channels can change the genetic and linguistic structure of populations. In traditional societies of eastern Indonesia, finely resolved cophylogenies of languages and genes reveal persistent movements between stable speech communities facilitated by kinship rules. When multiple languages are present in a region and postmarital residence rules encourage sustained directional movement between speech communities, then languages should be channeled along uniparental lines. We find strong evidence for this pattern in 982 individuals from 25 villages on two adjacent islands, where different kinship rules have been followed. Core groups of close relatives have stayed together for generations, while remaining in contact with, and marrying into, surrounding groups. Over time, these kinship systems shaped their gene and language phylogenies: Consistently following a postmarital residence rule turned social communities into speech communities.

60 APPLIED LIFE SCIENCES↗

Deep learning approaches for neural decoding across architectures and recording modalities

Decoding behavior, perception or cognitive state directly from neural signals is critical for brain–computer interface research and an important tool for systems neuroscience. In the last decade, deep learning has become the state-of-the-art method in many machine learning tasks ranging from speech recognition to image segmentation. The success of deep networks in other domains has led to a new wave of applications in neuroscience. In this article, we review deep learning approaches to neural decoding. Here, we describe the architectures used for extracting useful features from neural recording modalities ranging from spikes to functional magnetic resonance imaging. Furthermore, we explore how deep learning has been leveraged to predict common outputs including movement, speech and vision, with a focus on how pretrained deep networks can be incorporated as priors for complex decoding targets like acoustic speech or images. Deep learning has been shown to be a useful tool for improving the accuracy and flexibility of neural decoding across a wide range of tasks, and we point out areas for future scientific development.

59 BASIC BIOLOGICAL SCIENCES↗

Robotics control using isolated word recognition of voice input

A speech input/output system is presented that can be used to communicate with a task oriented system. Human speech commands and synthesized voice output extend conventional information exchange capabilities between man and machine by utilizing audio input and output channels. The speech input facility is comprised of a hardware feature extractor and a microprocessor implemented isolated word or phrase recognition system. The recognizer offers a medium sized (100 commands), syntactically constrained vocabulary, and exhibits close to real time performance. The major portion of the recognition processing required is accomplished through software, minimizing the complexity of the hardware feature extractor.

Weiner, J. M.↗

Biomedical technology transfer. Applications of NASA science and technology

Ongoing projects described address: (1) intracranial pressure monitoring; (2) versatile portable speech prosthesis; (3) cardiovascular magnetic measurements; (4) improved EMG biotelemetry for pediatrics; (5) ultrasonic kidney stone disintegration; (6) pediatric roentgen densitometry; (7) X-ray spatial frequency multiplexing; (8) mechanical impedance determination of bone strength; (9) visual-to-tactile mobility aid for the blind; (10) Purkinje image eyetracker and stabilized photocoalqulator; (11) neurological applications of NASA-SRI eyetracker; (12) ICU synthesized speech alarm; (13) NANOPHOR: microelectrophoresis instrument; (14) WRISTCOM: tactile communication system for the deaf-blind; (15) medical applications of NASA liquid-circulating garments; and (16) hip prosthesis with biotelemetry. Potential transfer projects include a person-portable versatile speech prosthesis, a critical care transport sytem, a clinical information system for cardiology, a programmable biofeedback orthosis for scoliosis a pediatric long-bone reconstruction, and spinal immobilization apparatus.

Harrison, D. C.↗

Noise-induced hearing impairment and handicap

A permanent, noise-induced hearing loss has doubly harmful effect on speech communications. First, the elevation in the threshold of hearing means that many speech sounds are too weak to be heard, and second, very intense speech sounds may appear to be distorted. The whole question of the impact of noise-induced hearing loss upon the impairments and handicaps experienced by people with such hearing losses was somewhat controversial partly because of the economic aspects of related practical noise control and workmen's compensation.

Source record↗

To speak or not to speak - A multiple resource perspective

The desirability of employing speech response in a dynamic dual task situation was discussed from a multiple resource perspective. A secondary task technique was employed to examine the time-sharing performance of five dual tasks with various degrees of resource overlap according to the structure-specific resource model of Wickens (1980). The primary task was a visual/manual tracking task which required spatial processing. The secondary task was either another tracking task or a spatial transformation task with one of four input (visual or auditory) and output (manual or speech) configurations. The results show that the dual task performance was best when the primary tracking task was paired with the visual/speech transformation task. This finding was explained by an interaction of the stimulus-central processing-response compatibility of the transformation task and the degree of resource competition between the time-shared tasks. Implications on the utility of speech response were discussed.

Tsang, P. S.↗

Response modalities and time-sharing performance

An experiment performed to investigate the role of resource competition and asymmetric transfer in dual-task performance is described. It is shown that there is an advantage to mixed manual/speech response modality configurations that cannot be accounted for by asymmetric transfer. The present results support the multiple resources approach to the application of speech technology. Once speech recognition achieves an acceptable level of operational reliability, speech controls can be used to reduce resource competition and improve performance in multitask environments.

Vidulich, Michael A.↗