Interpolation of LPC spectra via pole shifting

Goncharoff, Vladimir; Kaine-Krolak, M.

doi:10.1109/icassp.1995.479810

Cited by 10 publications

(5 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…High-quality speech stimuli were synthesized using an original morphing method based on a modified linear predictive coding (LPC) analysis synthesis scheme ( 64 ). Using an exponential pole-morphing approach, the second formant transition was morphed to build a linear speech sound continuum between /ba/, /da/, and /ga/ ( 65 , 66 ). The initial (prototypical) /ba/, /da/, and /ga/ syllables were natural voice signals, down-sampled to 16 kHz, aligned on their burst starting time, and cut to the same length (360 ms).…”

Section: Methodsmentioning

confidence: 99%

Focal versus distributed temporal cortex activity for speech sound category assignment

Bouton

Chambon

Tyrand

et al. 2018

Proc. Natl. Acad. Sci. U.S.A.

View full text Add to dashboard Cite

SignificanceWhen listening to speech, phonemes are represented in a distributed fashion in our temporal and prefrontal cortices. How these representations are selected in a phonemic decision context, and in particular whether distributed or focal neural information is required for explicit phoneme recognition, is unclear. We hypothesized that focal and early neural encoding of acoustic signals is sufficiently informative to access speech sound representations and permit phoneme recognition. We tested this hypothesis by combining a simple speech-phoneme categorization task with univariate and multivariate analyses of fMRI, magnetoencephalography, intracortical, and clinical data. We show that neural information available focally in the temporal cortex prior to decision-related neural activity is specific enough to account for human phonemic identification.

show abstract

Section: Methodsmentioning

confidence: 99%

Focal versus distributed temporal cortex activity for speech sound category assignment

Bouton

Chambon

Tyrand

et al. 2018

Proc. Natl. Acad. Sci. U.S.A.

View full text Add to dashboard Cite

show abstract

“…High-quality speech stimuli were synthesized using STRAIGHT (52) together with an original morphing method, based on a modified LPC (Linear Predictive Coding) analysis-synthesis scheme (53). The second formant transition, was morphed in order to build a linear speech sound continuum between /ba/, /da/ and /ga/ (54,55). Through the continuum, we kept constant a low-pass filtered pulse train for the voiced part, a filtered noise throughout all the stimulus, an additional white noise for the burst, the global amplitude of the stimulus, and the first and third formant transitions.…”

Section: Methodsmentioning

confidence: 99%

Focal versus distributed temporal cortex activity for speech sound category assignment

Bouton

Chambon

Tyrand

et al. 2017

Preprint

View full text Add to dashboard Cite

SummaryPercepts and words can be decoded from largely distributed neural activity measures. The existence of widespread representations might, however, conflict with the fundamental notions of hierarchical processing and efficient coding. Using fMRI and MEG during syllable identification, we first show that sensory and decisional activity co-localize to a restricted part of the posterior superior temporal cortex. Next, using intracortical recordings we demonstrate that early and focal neural activity in this region distinguishes correct from incorrect decisions and can be machine-decoded to classify syllables. Crucially, significant machine-decoding was possible from neuronal activity sampled across widespread regions, despite weak or absent sensory or decision-related responses. These findings show that a complex behavior like speech sound categorization relies on an efficient readout of focal neural activity, while distributed activity, although decodable by machine-learning, reflects collateral processes of sensory perception and decision.

show abstract

“…A metric that satisfies (12) can be said to be filtering invariant because of the following statistical interpretation. Any spectral density Φ with minimum-phase spectral factor W can be identified to a n-dimensional zero-mean second-order stationary purely nondeterministic stochastic process {y(t)} t∈Z generated by filtering a white noise process through W .…”

Section: A Filtering Invariancementioning

confidence: 99%

Conal Distances Between Rational Spectral Densities

Baggio

Ferrante

Sepulchre

2019

IEEE Trans. Automat. Contr.

View full text Add to dashboard Cite

The paper generalizes Thompson and Hilbert metric to the space of spectral densities. The resulting complete metric space has the differentiable structure of a Finsler manifold with explicit geodesics. The resulting distances are filtering invariant, can be computed efficiently, and admit geodesic paths that preserve rationality; these are properties of fundamental importance in many engineering applications.

show abstract

Interpolation of LPC spectra via pole shifting

Cited by 10 publications

References 0 publications

Focal versus distributed temporal cortex activity for speech sound category assignment

Focal versus distributed temporal cortex activity for speech sound category assignment

Focal versus distributed temporal cortex activity for speech sound category assignment

Conal Distances Between Rational Spectral Densities

Contact Info

Product

Resources

About