—Dictionary Learning (DL) techniques aim to find sparse signal representations that capture prominent characteristics in a given data. Such methods operate on a data matrix Y ∈ R N ×M , where each of its columns yi ∈ R N constitutes a training sample, and these columns together represent a sampling from the data manifold. For signals y ∈ R N residing on… (More)
—This work proposes a component based model for the raw ultrasound signals acquired by the transducer elements. Based on this approach, before undergoing the standard digital processing chain, every sampled raw signal is first decomposed into a smooth background signal and a strong reflectors component. The decomposition allows for a suited processing… (More)
In this paper, we propose a supervised dictionary learning algorithm that aims to preserve the local geometry in both dimensions of the data. A graph-based regularization explicitly takes into account the local manifold structure of the data points. A second graph regularization gives similar treatment to the feature domain and helps in learning a more… (More)
An autostereogram is a single image that encodes depth information that pops out when looking at it. The trick is achieved by setting a basic 2D pattern and continuously replicating the local pattern at each point in the image with a shift defined by the desired disparity. In this work, we explore the dependency between the ease of perceiving depth in… (More)
—In this work, we tackle the problem of multi-label classification using a sparsity-based approach. Multi-label classification problems, in which each instance is associated with a set of multiple labels, have received significant attention over the past few years due to the ongoing growth of data dimensions and availability. However, the dependency between… (More)
In this work, we propose a supervised dictionary learning algorithm , that attempts to preserve the local geometry in both dimensions of the data. A graph-based regularization explicitly takes into account the local manifold structure of the data, and a second graph regularization gives similar treatment to the feature domain and helps in learning a more… (More)
Given N points in the plane P1, P2, ..., PN and a location Ω, the union of discs with diameters [ΩPi], i = 1, 2, ..., N covers the convex hull of the points. The location Ωs minimizing the area covered by the union of discs, is shown to be the Steiner center of the convex hull of the points. Similar results for d-dimensional Euclidean space are conjectured.