News
Site News
Human hearing beats the Uncertainty Principle
Our psychophysical study of the acuity of simultaneous perception of the timing and frequency of a sound signal just appeared in Phys. Rev. Lett. We show that many of our subjects can make simultaneous judgements of the timing and frequency of a sound whose accuracies exceed the bounds from the Fourier Uncertainty Principle, sometimes by factors exceeding 10.
Thibaud is now officially Dr. Taillefumier.
On June 14 Thibaud Taillefumier attended Convocation. Here is a transcript of my remarks at the occasion.
Large-scale, exact integration of stochastic neural networks.
"Exact Event-Driven Implementation for Recurrent Networks of Stochastic Perfect Integrate-and-Fire Neurons", by Taillefumier, Touboul and Magnasco, in press at Neural Computation.
Blind deconvolution algorithm for electrophysiological potentials.
We present a sophisticated algorithm for the analysis of excitatory postsynaptic currents, capable of resolving overlapping sums of such EPSCs to 0.25 millisecond resolution.
The Veins of a Leaf: Revealing Nature's Mathematical System
A newsfeature in NSF's Science Nation, picked up by PBS Newshour
The Sound of Running Water
Auditory textures such as the sound of running water, rain, fire, wind, or insects are readily recognized by our auditory system. We study running water and showed its sound has many interesting physical features, such as scale invariance. We use these features to create artificially synthesized versions of such sounds. We study human listeners' perception of these sounds, and what features in the sound lead to its categorization as a "natural" sound.
Self-Tuned Critical Anti-Hebbian Networks
In this paper we present an abstract model for the "operating point" of the brain. We hypothesize that the brain poises itself at an "extensive" critical point, i.e., one in whch there is a large number of degrees of freedom poised precisely at the onset of an oscillatory instability. This dynamical criticality leads to states that are globally coupled. Published online in Physical Review Letters: http://link.aps.org/doi/10.1103/PhysRevLett.102.258102
A Haar-like basis for the Ornstein-Uhlenbeck process.
A new recursive way to construct brownian motion.

