Massachusetts Institute of Technology

Christopher J Smalt

Thomas Francis Quatieri, Jr

Trina Rae Vian

Joseph John Lacirignola

David Christopher Maurer

David Francis Aubin, Jr

Paula P Collins

Nicolas Malyska

Kate D Fischl

Paul D Gatewood

Gregory Alan Ciccarelli

Paul T Calamia

Edward H Chen

Kerry Alexander Johnson

Nima Mesgarani

James O'Sullivan

Christine Michelle Weston

Michael S Brandstein

Satrajit S Ghosh

James R Williamson

Jeffrey Shane Palmer

Joseph Perricone

Michael Nolan

Mary Katherine Byrd

Stephanie Haro

Shakti Davis

Olha Townsend

In one aspect of the present disclosure, method includes: receiving neural data responsive to a listener's auditory attention; receiving an acoustic signal responsive to a plurality of acoustic sources; for each of the plurality of acoustic sources: generating, from the received acoustic signal, audio data comprising one or more features of the acoustic source, forming combined data representative of the neural data and the audio data, and providing the combined data to a classification network configured to calculate a similarity score between the neural data and the acoustic source using one or more similarity metrics; and using the similarity scores calculated for each of the acoustic sources to identify, from the plurality of acoustic sources, an acoustic source associated with the listener's auditory attention.

End-to-end deep neural network for auditory attention decoding

Systems, apparatus, and methods for collecting, interpreting, and utilizing noise exposure data may include sensors to obtain an analog signal representative of impulse noise sound pressure and an analog signal representative of continuous noise sound pressure. At least one ADC may generate digital signals by sampling the analog signals at rates equal to or greater than twice the reciprocal of a minimum impulse noise rise time. Accelerometers may obtain data in close proximity to and remote from the sensors. At least one processor may include a first combining node to combine the digital signals to represent both the continuous noise and the impulse noise, a shock-artifact detection filter to identify a time frame including a shock artifact based on the accelerometry data, a frequency filter to generate a background-removed audio signal, an adaptive filter to estimate the shock artifact, and a second combining node to produce a shock-artifact-free audio signal.

High fidelity systems, apparatus, and methods for collecting noise exposure data

Loud sounds with fast rise times, like gunfire and explosions, can cause noise-induced hearing loss (NIHL). Unfortunately, current models do not adequately explain how impulsive sounds cause NIHL, which makes it difficult to predict and prevent NIHL on battlefields and other hostile or rugged environments. Fortunately, the impulsive sounds experienced by soldiers and others working in rugged environments can be recorded using a compact, portable system that acquires, digitizes, and stores high-bandwidth audio data. An example of this system can be mounted on a helmet or other article and used to record hours of audio data at a bandwidth of 20 kHz or higher, which is broad enough to capture sounds with rise times less than 50 ms. An analog-to-digital converter (ADC) digitizes these broadband audio signals at rate of 40 kHz or higher to preserve the impulse information. A processor transfers the digitized samples from a buffer to a memory card for later retrieval using an interrupt-driven processing technique.

Methods and apparatus for recording impulsive sounds

In a system and method for assessing the condition of a subject, control parameters are derived from a neurophysiological computational model that operates on features extracted from a speech signal. The control parameters are used as biomarkers (indicators) of the subject's condition. Speech related features are compared with model predicted speech features, and the error signal is used to update control parameters within the neurophysiological computational model. The updated control parameters are processed in a comparison with parameters associated with the disorder in a library.

Assessing disorders through speech and a computational model

Growing community of inventors

Arlington, MA, United States of America

Christopher J Smalt