A new method for self-supervised sensorimotor learning of sound source localization is presented, that allows a simulated listener to learn online an auditorimotor map from the sensorimotor experience provided by an auditory evoked behavior. The map represents the auditory space and is used to estimate the azimuthal direction of sound sources. The learning mainly consists in non-linear dimensionality reduction of sensorimotor data. Our results show that an auditorimotor map can be learned, both from real and simulated data, and that the online learning leads to accurate estimations of azimuthal sources direction.

Newborns must learn to structure incoming acoustic information into segments, words, phrases, etc., before they can start to learn language. This process is thought to rely on modulation structure of the speech waveform induced by segmental or prosodic regularities within the speech heard by the infant. Here, we investigate the process by which the initial acoustic processing required by modulation analysis can itself be tuned by exposure to the regularities of speech. Starting from the classic definition of modulation, as applied within channels of the peripheral filter, we formulate a mathematical framework in which the structure of initial spectral filtering is adapted for modulation analysis. Our working hypothesis is that the human ear and brain are adapted to the analysis of modulation, via a data-driven learning process on the scale of development (or possibly evolution). Simulation results are presented and a comparison with filterbanks classically used in signal processing is done.

How does one extract unknown but stereotypical events that are linearly superimposed within a signal with variable latencies and variable amplitudes? One could think of using template matching or matching pursuit to find the arbitrarily shifted linear components. However, traditional matching approaches require that the templates be known a priori. To overcome this restriction we use instead semi Non-Negative Matrix Factorization (semi-NMF) that we extend to allow for time shifts when matching the templates to the signal. The algorithm estimates templates directly from the data along with their non-negative amplitudes. The resulting method can be thought of as an adaptive template matching procedure. We demonstrate the procedure on the task of extracting spikes from single channel extracellular recordings. On these data the algorithm essentially performs spike detection and unsupervised spike clustering. Results on simulated data and extracellular recordings indicate that the method performs well for signal-to-noise ratios of 6dB or higher and that spike templates are recovered accurately provided they are sufficiently different.

We present a method for removing environmental noise from physiological recordings such as magnetoencephalography (MEG) for which noise-sensitive reference channels are available. Sensor signals are projected on a subspace spanned by the reference channels augmented by time-shifted and/or nonlinearly transformed versions of the same, and the projections are removed to obtain "clean" sensor signals. The method compensates for scalar, convolutional or non-linear mismatches between sensor and reference channels by synthesizing, for each reference/sensor pair, a filter that is optimal in a least-squares sense for removal of the artifact. The method was tested with synthetic and real MEG data, typically removing up to 98% of noise variance. It offers an alternative to bulky and costly magnetic shielding (multiple layers of aluminium and mu-metal) for present scientific and medical applications and future developments such as brain-machine interfaces (BMI).

We present in this paper a novel F0contour estimation method based on a parametric description of the wavelet power spectrum of speech that accounts for its structure simultaneously in time and frequency directions. We model the speech spectrum as a sequence of spectral clusters governed by a smooth common F0contour expressed as a spline curve. The harmonic and temporal structure of these clusters and their common F0contour are estimated simultaneously. Through experimental comparisons with existing methods, we show that our algorithm is competitive on clean single-speaker speech, and that it outperforms existing methods both in the presence of noise and for the estimation of multiple F0contours of cochannel concurrent speech.

This paper proposes a novel $F_{0}$ contour estimation algorithm based on a precise parametric description of the voiced parts of speech derived from the power spectrum. The algorithm is able to perform in a wide variety of noisy environments as well as to estimate the $F_{0}$ s of cochannel concurrent speech. The speech spectrum is modeled as a sequence of spectral clusters governed by a common $F_{0}$ contour expressed as a spline curve. These clusters are obtained by an unsupervised 2-D time-frequency clustering of the power density using a new formulation of the EM algorithm, and their common $F_{0}$ contour is estimated at the same time. A smooth $F_{0}$ contour is extracted for the whole utterance, linking together its voiced parts. A noise model is used to cope with nonharmonic background noise, which would otherwise interfere with the clustering of the harmonic portions of speech. We evaluate our algorithm in comparison with existing methods on several tasks, and show 1) that it is competitive on clean single-speaker speech, 2) that it outperforms existing methods in the presence of noise, and 3) that it outperforms existing methods for the estimation of multiple $F_{0}$ contours of cochannel concurrent speech.

A method is proposed for the analysis of acoustic scenes. The contribution of each competing source is suppressed on the basis of harmonic structure or cross-sensor correlation, in such a way that other sources may be estimated. Successive suppression of sources allows the scene to be characterized. In the limit of purely periodic sources and no noise, the method provides accurate estimates of fundamental frequency and spectral content. In the presence of noise or imperfect periodicity, it offers likelihood functions from which the spectral content of sources may be inferred using Bayesian methods. The method is an alternative to more familiar spectral and spectro-temporal methods of acoustic scene analysis. An advantage is that precise analysis is possible using short temporal windows.

A methodology is presented for fundamental frequency estimation of one or more voices. The signal is modeled as the sum of one or more periodic signals, and the parameters estimated by search with interpolation. Accurate, reliable estimates are obtained for each frame without tracking or continuity constraints, and without the use of specific instrument models (although their use might further boost performance). In formal evaluation over a large database of speech, the single-voice algorithm outperformed the best competing methods by a factor of three. 1.

cote interne IRCAM: Kawahara02a

This paper reports the comparative evaluation of several speech F0 evaluation algorithms over a wide database of laryngographlabeled speech. Included are several classic algorithms that are available in software on the net, as well as two new algorithms that offer greatly reduced error rates. Particular attention is given to the methodology of evaluation.

cote interne IRCAM: DeCheveigne94c

This mixed speech fundamental frequency (f 0 ) estimation algorithm is an extension of the classical AMDF (Average Magnitude Difference Function) algorithm for one voice. An exhaustive search of the parameter space of two cascaded timedomain comb filters yields an estimation of the periods of the component voices. The algorithm, which is computationally expensive but easily parallelizable, was tested on a database of continuous male and female speech. Segments of voiced speech, selected according to a "good periodicity" criterion to ensure that the reference single-voice f 0 algorithm would not fail (this criterion rejected 25% of voiced speech frames), were paired and summed to simulate mixed speech. The search range of the algorithm was limited to a 3 octave range, and search was performed frame-by-frame without continuity constraints. The resulting estimates were compared to those of the reference algorithm and found to be within 3 % of target values for 90 % of all frames. Keyword...

Recently published data argue in favor of the hypothesis of a time-domain code in the representation of sound in the auditory nerve. This paper describes a pitch perception model based on such a code, and relates it's behaviour to pitch theory. The model accounts for the pitch of complex tones, and correctly predicts pitch shifts and phase effects. It is physiologically realistic.