All Years Seminars
[INMA] 2016-04-19 (14:00) : Saccadic mask inhibition: from "it-has-to-be-a-collicular-mechanism" to the impossible color quest
At Euler building (A002)
Speaker :
Daye, Pierre
Abstract :
This presentation will be divided into two parts. First, I will present a robust mechanism called "saccadic mask inhibition" induced by displaying a distractor a short period after a saccadic target has been presented. The distractor delays the initiation of a large proportion of saccades but has little effect on landing positions. Physiologically, it is postulated that the saccadic mask inhibition results from a competition between the target and the distractor in the visuomotor maps of the superior colliculus without any cortical involvement. To test this assumption, we updated the original version of the protocol to include conditions in which the collicular involvement is disputable. Because of the nature of the updated protocol, we had to create a new test to evaluate the color perception of each subject. For more than three decades, researchers have worked on a protocol to evaluate the perception of color isoluminance. One of the major issues of the current protocols is that they require a response of the subjects (e.g. "this color is more lighter/darker than this one"). Aside from the obvious bias that could be induced by an answer of the subjects, these protocols could not be used on non-cooperative subjects (infants, patients, monkeys, etc). We developed a new protocol that can be used to evaluate the isoluminance between two colors without requiring a perceptual answer of the subject. Using a combination of the two tests, we were able to demonstrate that the saccadic mask inhibition mechanism does not originate from the superior colliculus but involves cortical mechanisms.
[INMA] 2016-04-12 (14:00) : Transfer Entropy and Directed Information in Gaussian Diffusion Processes
At Euler building (A002)
Speaker :
Nigel J. Newton (University of Essex)
Abstract :
Transfer Entropy and Directed Information are information-theoretic measures of the directional dependency between stochastic processes. They were both originally defined for discrete-time processes: Directed Information in the context of telecommunication systems with feedback, and Transfer Entropy as a means of testing for directional influence in complex systems. The talk will define and evaluate the Transfer Entropy and Directed Information between components of multidimensional Gaussian diffusion processes. When the components are jointly Markov, the natural definitions give rise to measures of influence according to a simple physical principle. More generally, the effect of other components has to be accounted for, and this can be achieved in more than one way. Both quantities can be expressed in terms of the solutions of matrix Riccati equations, and so are easy to compute. Applications to Nonequilibrium Statistical Mechanics will be briefly discussed.
[INMA] 2016-04-06 (14:00) : Fast reconstruction of a compressively sampled data stream
At Euler building (A002)
Speaker :
Pantelis Sopasakis (IMT Institute for Advanced Studies Lucca)
Abstract :
A very wide spectrum of optimization problems can be efficiently solved with proximal gradient methods which hinge on the celebrated forward-backward splitting (FBS) schema. But such first-order methods are only effective when low or medium accuracy is required and are known to be rather slow or even impractical for badly conditioned problems. Moreover, the straightforward introduction of second-order (Hessian) information is beset with shortcomings as, typically, at every iteration we need to solve a non-separable optimisation problem.
In this talk we will follow a different route to the solution of such optimisation problems. We will recast non-smooth optimisation problems as the minimisation of a real-valued, continuously differentiable function known as the forward-backward envelope. We will then employ a semismooth Newton method to solve the equivalent optimisation problem instead of the original one. We will then apply the proposed semismooth Newton method to L1-regularised least squares (LASSO) problems which is motivated by an an interesting application: recursive compressed sensing.
Compressed sensing is a signal processing methodology for the reconstruction of sparsely sampled signals and it offers a new paradigm for sampling signals based on their innovation, that is, the minimum number of coefficients sufficient to accurately represent it in an appropriately selected basis. Compressed sensing leads to a lower sampling rate compared to theories using some fixed basis and has many applications in image processing, medical imaging and MRI, photography, holography, facial recognition, radio astronomy, radar technology and more. The traditional compressed sensing approach is naturally offline, in that it amounts to sparsely sampling and reconstructing a given dataset. Recently, an online algorithm for performing compressed sensing on streaming data was proposed; the scheme uses recursive sampling of the input stream and recursive decompression to accurately estimate stream entries from the acquired noisy measurements.
We will see how we can tailor the forward-backward Newton method to solve recursive compressed sensing problems at one tenth of the time required by other algorithms such as ISTA, FISTA, ADMM and interior-point methods (L1LS).
[INMA] 2016-03-15 (14:00) : Cell Segmentation with Random Ferns and Graph Cuts
At Euler building (A002)
Speaker :
Browet, Arnaud
Abstract :
The progress in imaging techniques have allowed the study of various aspect of cellular mechanisms. To isolate individual cells in live imaging data, we introduce an elegant image segmentation framework that effectively extracts cell boundaries, even in the presence of poor edge details. Our approach works in two stages. First we estimate pixel interior/border/exterior class probabilities using a random ferns classifier. Then we use an energy minimization framework to compute boundaries whose localization is compliant with the pixel class probabilities. We validate our approach on a manually annotated dataset and demonstrate that our segmentation procedure is in very good agreement with the ground-truth defined by biologists.
[INMA] 2016-03-08 (14:00) : The ground truth about metadata and community detection in networks
At Euler building (A002)
Speaker :
Peel, Leto
Abstract :
It is common to evaluate the performance of community detection algorithms by
their ability to find so-called ground truth communities. This works well in synthetic
networks with planted partitions because such networks' links are formed explicitly based
on the planted partition. However, there are no planted communities in real world networks,
so in their place, it is common to use discrete-valued node attributes, or metadata,
to define a partition to act as ground truth. In this work, we first argue that metadata are
not the same as ground truth, and that to treat them as such raises theoretical and practical
concerns. Nevertheless, understanding the relationship between metadata and community structure
is important, so we subsequently introduce two new techniques to rigorously explore this relationship.
Seminars
Subscribe (ICS)
INMA Contact Info
Mathematical Engineering (INMA)
L4.05.01
Avenue Georges Lemaître, 4
+32 10 47 80 36
secretaire-inma@uclouvain.be
Mon – Fri 9:00A.M. – 5:00P.M.
JOIN US