A5055470679- Phonetics and Phonology Research
- Speech Recognition and Synthesis
- Linguistic Variation and Morphology
- Speech and Audio Processing
- EEG and Brain-Computer Interfaces
- Music and Audio Processing
- Language and cultural evolution
- Neurobiology of Language and Bilingualism
- Sleep and Wakefulness Research
- Speech and dialogue systems
- Sleep and related disorders
- Natural Language Processing Techniques
- Reading and Literacy Development
- Neuroscience and Music Perception
- Obstructive Sleep Apnea Research
- Multisensory perception and integration
- Child and Animal Learning Development
- Time Series Analysis and Forecasting
- Music Technology and Sound Studies
- Intelligent Tutoring Systems and Adaptive Learning
- Language Development and Disorders
- Machine Learning and Algorithms
- Memory Processes and Influences
- Advanced Memory and Neural Computing
- Numerical Methods and Algorithms
Massachusetts General Hospital
2022
Rutgers Sexual and Reproductive Health and Rights
2019-2020
Rutgers, The State University of New Jersey
2019
University of Rochester
2011-2018
Neuroscience Institute
2018
Princeton University
2017-2018
One of the persistent puzzles in understanding human speech perception is how listeners cope with talker variability. thing that might help structure variability: rather than varying randomly, talkers same gender, dialect, age, etc. tend to produce language similar ways. Listeners are sensitive this covariation between linguistic variation and socio-indexical variables. In paper I present new techniques based on ideal observer models quantify (1) amount type (informativity a grouping...
We present a framework of second and additional language (L2/L n ) acquisition motivated by recent work on socio‐indexical knowledge in first (L1) processing. The distribution linguistic categories covaries with variables (e.g., talker identity, gender, dialects). summarize evidence that implicit probabilistic this covariance is critical to L1 processing, propose L2/L learning uses the same type information probabilistically infer latent hierarchical structure over previously learned new...
Brain imaging researchers regularly work with large, heterogeneous, high-dimensional datasets.Historically, have dealt this complexity idiosyncratically, every lab or individual implementing their own preprocessing and analysis procedures.The resulting lack of field-wide standards has severely limited reproducibility data sharing reuse.
Abstract Social and linguistic perceptions are linked. On one hand, talker identity affects speech perception. the other itself provides information about a talker's identity. Here, we propose that same probabilistic knowledge might underlie both socially conditioned inferences linguistically social inferences. Our computational–level approach—the ideal adapter—starts from idea listeners use of covariation between social, linguistic, acoustic cues in order to infer most likely explanation...
In their article “Evolved structure of language shows lineage-specific trendsin word order universals”, Dunn, Greenhill, Levinson, & Gray present evi-dence purporting to demonstrate that both Chomskyan and Greenbergian lan-guage universals are invalid. particular, most interest readers ofthis journal, they state “contrary the generalizations, we showthat observed functional dependencies between traits lineage-specificrather than universal tendencies” (Dunn et al. 2011: 79). If this...
Abstract A listener's interpretation of a given speech sound can vary probabilistically from moment to moment. Previous experience (i.e., the contexts in which one has encountered an ambiguous sound) further influence speech, phenomenon known as perceptual learning for speech. This study used multivoxel pattern analysis query how neural patterns reflect learning, leveraging archival fMRI data lexically guided conducted by Myers and Mesite [Myers, E. B., & Mesite, L. M. Neural systems...
Perceptual systems have to make sense out of a world that is not only noisy and ambiguous, but also varies from situation situation. Human speech perception perceptual domain where this problem has long been acknowledged: individual talkers vary substantially in how they produce linguistic units using acoustic cues. Yet, the system solves talker variability remains poorly understood. This thesis presents computational framework---the ideal adapter---for understanding it. The basic insight...
Abstract Introduction Accurate sleep staging of EEG data from polysomnography (PSG) is important in the diagnosis narcolepsy. Human costly and labor intensive, but automated algorithms must be rigorously tested narcoleptic patients to ensure valid performance. PSGs often tend more fragmented variable than non-narcoleptic populations, making it challenging for both humans accurately stage sleep. Here, we evaluate performance a deep learning model validated general clinic population nocturnal...
Abstract Introduction Automated algorithms for assisting sleep technologists and clinicians in the staging of have potential to significantly speed scoring PSGs, reduce inter-rater variability, improve reliability diagnosis disorders. Here, we describe an automated machine learning algorithm that scores PSG using only EEG signals. Methods We a convolutional neural network-based algorithm, SleepStageML™, was trained on database >19,000 polysomnography recordings from heterogeneous...
Abstract Introduction Sleep monitoring hardware that allows for accurate sleep staging while also being unobtrusive and self-administered has the potential to make reliable EEG-based assessment of quality at home widely accessible. However, novel utilizing dry EEG electrodes, a question remains regarding similarity these signals traditional polysomnography (PSG), especially when used therapy development. Here, we investigate how well from wearable electrode system can be staged by an...
Abstract Background Sleep spindle activity is commonly estimated by measuring sigma power during stage 2 non-rapid eye movement (NREM2) sleep. However, spindles account for little of the total NREM2 interval and therefore over entire may be misleading. This study compares derived measures from direct automated detection with those gross spectral analyses purposes clinical trial design. Methods We in a set 8,440 overnight electroencephalogram (EEG) recordings 5,793 patients Heart Health Study...
In order to parse speech in real time, listeners should use any informative cues available. Here, we investigate the role of segmental duration. Previous work has found statistically significant differences mean durations analogous segments across different lexical/syntactic structures. However, a difference means does not necessarily that distributions these make individual token sufficiently be useful cue. The goal this is production data quantify how duration about syntactic/lexical...