Searcharxiv⌕ Search

arXiv subjects

Zoltan Szabo

Publications and source records attributed to Zoltan Szabo.

At least 55 records · Page 3Linked to original sources
arXiv2013

Emotional Expression Classification using Time-Series Kernels

Estimation of facial expressions, as spatio-temporal processes, can take advantage of kernel methods if one considers facial landmark positions and their motion in 3D space. We applied support vector classification with kernels derived from dynamic time-warping similarity measures. We achieved over 99% accuracy - measured by area under ROC curve - using only the 'motion pattern' of the PCA compressed representation of the marker point vector, the so-called shape parameters. Beyond the classification of full motion patterns, several expressions were recognized with over 90% accuracy in as few as 5-6 frames from their onset, about 200 milliseconds.

cs.CV↗
arXiv2012

Distributed High Dimensional Information Theoretical Image Registration via Random Projections

Information theoretical measures, such as entropy, mutual information, and various divergences, exhibit robust characteristics in image registration applications. However, the estimation of these quantities is computationally intensive in high dimensions. On the other hand, consistent estimation from pairwise distances of the sample points is possible, which suits random projection (RP) based low dimensional embeddings. We adapt the RP technique to this task by means of a simple ensemble method. To the best of our knowledge, this is the first distributed, RP based information theoretical image registration approach. The efficiency of the method is demonstrated through numerical examples.

cs.IT↗
arXiv2012

Automated Word Puzzle Generation via Topic Dictionaries

We propose a general method for automated word puzzle generation. Contrary to previous approaches in this novel field, the presented method does not rely on highly structured datasets obtained with serious human annotation effort: it only needs an unstructured and unannotated corpus (i.e., document collection) as input. The method builds upon two additional pillars: (i) a topic model, which induces a topic dictionary from the input corpus (examples include e.g., latent semantic analysis, group-structured dictionaries or latent Dirichlet allocation), and (ii) a semantic similarity measure of word pairs. Our method can (i) generate automatically a large number of proper word puzzles of different types, including the odd one out, choose the related word and separate the topics puzzle. (ii) It can easily create domain-specific puzzles by replacing the corpus component. (iii) It is also capable of automatically generating puzzles with parameterizable levels of difficulty suitable for, e.g., beginners or intermediate learners.

cs.CL↗
arXiv2012

Combinatorial Heegaard Floer homology and nice Heegaard diagrams

We consider a stabilized version of hat Heegaard Floer homology of a 3-manifold Y (i.e. the U=0 variant of Heegaard Floer homology for closed 3-manifolds). We give a combinatorial algorithm for constructing this invariant, starting from a Heegaard decomposition for Y, and give a combinatorial proof of its invariance properties.

math.GT↗
arXiv2012

Collaborative Filtering via Group-Structured Dictionary Learning

Structured sparse coding and the related structured dictionary learning problems are novel research areas in machine learning. In this paper we present a new application of structured dictionary learning for collaborative filtering based recommender systems. Our extensive numerical experiments demonstrate that the presented technique outperforms its state-of-the-art competitors and has several advantages over approaches that do not put structured constraints on the dictionary elements.

math.OC↗
arXiv2010

Towards Nonstationary, Nonparametric Independent Process Analysis with Unknown Source Component Dimensions

The goal of this paper is to extend independent subspace analysis (ISA) to the case of (i) nonparametric, not strictly stationary source dynamics and (ii) unknown source component dimensions. We make use of functional autoregressive (fAR) processes to model the temporal evolution of the hidden sources. An extension of the ISA separation principle--which states that the ISA problem can be solved by traditional independent component analysis (ICA) and clustering of the ICA elements--is derived for the solution of the defined fAR independent process analysis task (fAR-IPA): applying fAR identification we reduce the problem to ISA. A local averaging approach, the Nadaraya-Watson kernel regression technique is adapted to obtain strongly consistent fAR estimation. We extend the Amari-index to different dimensional components and illustrate the efficiency of the fAR-IPA approach by numerical examples.

stat.ME↗
arXiv2010

On combinatorial link Floer homology

Link Floer homology is an invariant for links defined using a suitable version of Lagrangian Floer homology. In an earlier paper, this invariant was given a combinatorial description with mod 2 coefficients. In the present paper, we give a self-contained presentation of the basic properties of link Floer homology, including an elementary proof of its invariance. We also fix signs for the differentials, so that the theory is defined with integer coefficients.

math.GT↗
arXiv2009

Floer homology and singular knots

We define Floer homology theories for oriented, singular knots in S^3 and show that one of these theories can be defined combinatorially for planar singular knots.

math.GT↗
arXiv2008

Rational blow-downs and smoothings of surface singularities

In this paper we give a necessary combinatorial condition for a negative--definite plumbing tree to be suitable for rational blow--down, or to be the graph of a complex surface singularity which admits a rational homology disk smoothing. New examples of surface singularities with rational homology disk smoothings are also presented; these include singularities with resolution graph having valency four nodes.

math.GT↗
arXiv2007

Link Floer homology and the Thurston norm

We show that link Floer homology detects the Thurston norm of a link complement. As an application, we show that the Thurston polytope of an alternating link is dual to the Newton polytope of its multi-variable Alexander polynomial. To illustrate these techniques, we also compute the Thurston polytopes of several specific link complements.

math.GT↗
arXiv2007

Knot Floer homology and integer surgeries

Let $K$ be a null-homologous knot in a three-manifold $Y$. We give a description of the Heegaard Floer homology of integer surgeries on $Y$ along $K$ in terms of the filtered homotopy type of the knot invariant for $K$. As an illustration, we calculate the Heegaard Floer homology groups of non-trivial circle bundles over Riemann surfaces.

math.GT↗
arXiv2007

Odd Khovanov homology

We describe an invariant of links in the three-sphere which is closely related to Khovanov's Jones polynomial homology. Our construction replaces the symmetric algebra appearing in Khovanov's definition with an exterior algebra. The two invariants have the same reduction modulo 2, but differ over the rationals. There is a reduced version which is a link invariant whose graded Euler characteristic is the normalized Jones polynomial.

math.QA↗
arXiv2007

On the skein exact squence for knot Floer homology

The aim of this paper is to study the skein exact sequence for knot Floer homology. We prove precise graded version of this sequence, and also one using $\HFm$. Moreover, a complete argument is also given purely within the realm of grid diagrams.

math.GT↗
arXiv2007

Undercomplete Blind Subspace Deconvolution via Linear Prediction

We present a novel solution technique for the blind subspace deconvolution (BSSD) problem, where temporal convolution of multidimensional hidden independent components is observed and the task is to uncover the hidden components using the observation only. We carry out this task for the undercomplete case (uBSSD): we reduce the original uBSSD task via linear prediction to independent subspace analysis (ISA), which we can solve. As it has been shown recently, applying temporal concatenation can also reduce uBSSD to ISA, but the associated ISA problem can easily become `high dimensional' [1]. The new reduction method circumvents this dimensionality problem. We perform detailed studies on the efficiency of the proposed technique by means of numerical simulations. We have found several advantages: our method can achieve high quality estimations for smaller number of samples and it can cope with deeper temporal convolutions.

stat.ME↗
arXiv2007

A cube of resolutions for knot Floer homology

We develop a skein exact sequence for knot Floer homology, involving singular knots. This leads to an explicit, algebraic description of knot Floer homology in terms of a braid projection of the knot.

math.GT↗