Xavier Siebert

• Associate Professor
• University of Mons

The growing demand for advanced tools to ensure safety in railway construction projects highlights the need for systems that can smoothly integrate and analyze multiple data modalities, such as multimodal learning algorithms. The latter, inspired by the human brain’s ability to integrate many sensory inputs, has emerged as a promising field in arti...

The rapid growth of e-commerce has placed considerable pressure on customs representatives, prompting advanced methods. In tackling this, Artificial intelligence (AI) systems have emerged as a promising approach to minimize the risks faced. Given that the Harmonized System (HS) code is a crucial element for an accurate customs declaration, we propo...

Building a prediction model based on both real data and expert knowledge can be challenging. We experienced this difficulty in our context of predicting fall risk factors in the elderly, where two sources of data were available: hand- reported information from medical consultations and expert information from human specialists. As is often the case...

In the era of artificial intelligence (AI), the deployment of intelligent systems for autonomous decision making has surged across diverse fields. However, the widespread adoption of AI technology is hindered by the risks associated with ceding control to autonomous systems, particularly in critical domains. Explainable artificial intelligence (XAI...

Active learning is a paradigm of machine learning which aims at reducing the amount of labeled data needed to train a classifier. Its overall principle is to sequentially select the most informative data points, which amounts to determining the uncertainty of regions of the input space. The main challenge lies in building a procedure that is comput...

The evaluation of risk factors for falls (RFF) is a key point in fall prevention for the elderly. Since the information of the main actionable RFF can not always be regularly re-evaluated by medical factors, their automatic prediction would allow providing useful recommendations to reduce the risk of falls. This article explores the advantages of t...

Active learning is a paradigm of machine learning which aims at reducing the amount of labeled data needed to train a classifier. Its overall principle is to sequentially select the most informative data points, which amounts to determining the uncertainty of regions of the input space. The main challenge lies in building a procedure that is comput...

Prevention of falls requires providing a small number of recommendations based on the risk factors present for a person. This article deals with the evaluation of 12 modifiable risk factors for fall, based on a selection of 45 variables from a real data set. The results of four classifiers (Logistic Regression, Random Forest, Artificial Neural Netw...

Constrained low-rank matrix approximations have been known for decades as powerful linear dimensionality reduction techniques able to extract the information contained in large data sets in a relevant way. However, such low-rank approaches are unable to mine complex, interleaved features that underlie hierarchical semantics. Recently, deep matrix f...

Background: Falls in the elderly are the number one cause of traumatic death in this population. Prevention of falls requires to evaluate which risk factors for fall are present for a person on the basis of available health information. Our objective is to predict the presence or the absence of 12 risk factors for fall in elderly people based on pa...

Active learning is typically used to label data, when the labeling process is expensive. Several active learning algorithms have been theoretically proved to perform better than their passive counterpart. However, these algorithms rely on some assumptions, which themselves contain some specific parameters. This paper adresses the problem of adaptiv...

Abstract—Falls are frequent in the elderly and it is the number one cause of traumatic death in this population. Fall is caused by many factors such as age-related physical changes or clinical factors, or the environment. Fall prevention requires to evaluate which risk factors for fall are present for a person on the basis of incomplete health data...

Constrained low-rank matrix approximations have been known for decades as powerful linear dimensionality reduction techniques to be able to extract the information contained in large data sets in a relevant way. However, such low-rank approaches are unable to mine complex, interleaved features that underlie hierarchical semantics. Recently, deep ma...

Breast cancer screening is a public health issue. Knowing that one in nine women will be affected by this disease, the detection of its first signs is crucial. Computer Assisted Diagnostic Systems (CADx) can help the radiologist to read mammograms and play a key role in the early detection of breast cancer. Nowadays, machine learning algorithms are...

Nonnegative matrix factorization (NMF) is a widely used linear dimensionality reduction technique for nonnegative data. NMF requires that each data point is approximated by a convex combination of basis elements. Archetypal analysis (AA), also referred to as convex NMF, is a well-known NMF variant imposing that the basis elements are themselves con...

Nonnegative matrix factorization (NMF) is a widely used linear dimensionality reduction technique for nonnegative data. NMF requires that each data point is approximated by a convex combination of basis elements. Archetypal analysis (AA), also referred to as convex NMF, is a well-known NMF variant imposing that the basis elements are themselves con...

Abstract Audio source separation concerns techniques used to extract unknown signals called sources from a mixed signal. In this paper, we assume that the audio signal is recorded with a single microphone. Considering a mixed signal composed of various audio sources, the blind audio source separation consists in isolating and extracting each of the...

There is a large body of work on convergence rates either in passive or active learning. Here we outline some of the results that have been obtained, more specifically in a nonparametric setting under assumptions about the smoothness and the margin noise. We also discuss the relative merits of these underlying assumptions by putting active learning...

Missing data cause various problems when processing and analysing real-world datasets. In this paper, we consider the structure of data from social networks’ accounts and introduce models of imputation and their ensembles designed for this area. We have analysed the structure and strength of correlations between data from social networks’ accounts...

Sound designers source sounds in massive collections, heavily tagged by themselves and sound librarians. For each query, once successive keywords attained a limit to filter down the results, hundreds of sounds are left to be reviewed. AudioMetro combines a new content-based information visualization technique with instant audio feedback to facilita...

This work focuses on label ranking, a particular task of preference learning, wherein the problem is to learn a mapping from instances to rankings over a finite set of labels. This paper discusses and proposes alternative reduction techniques that decompose the original problem into binary classification related to pairs of labels and that can take...

VideoCycle is a candidate application for this second Video Browser Showdown challenge. VideoCycle allows interactive intra-video and inter-shot navigation with dedicated gestural controllers. MediaCycle, the framework it is built upon, provides media organization by similarity, with a modular architecture enabling most of its workflow to be perfor...

This book constitutes the proceedings of the 5th International Conference on Intelligent Technologies for Interactive Entertainment, INTETAIN 2013. The 23 full papers presented were carefully selected from numerous submissions. The conference aims at enhancing the understanding of recent and anticipated advances in interactive technologies, and the...

This work focuses on a particular application of preference ranking, wherein the problem is to learn a mapping from instances to rankings over a finite set of labels, i.e. label ranking. Our approach is based on a learning reduction technique and provides such a mapping in the form of logical rules: if [antecedent] then [consequent], where [anteced...

Even though video compression has become a mature field, a lot of research is still ongoing. Indeed, as the quality of the compressed video for a given size or bit rate increases, so does users’ level of expectations and their intolerance to artefacts. The development of compression technology has enabled number of applications; key applications in...

Peptidylglycine α-hydroxylating monooxygenase (PHM) catalyzes the stereospecific hydroxylation of the Cα of C-terminal glycine-extended peptides and proteins, the first step in the activation of many peptide hormones, growth factors, and neurotransmitters. The crystal structure of the enzyme revealed two nonequivalent Cu sites (Cu(M) and Cu(H)) sep...

In this paper, we present an interactive installation allowing to navigate through a collection of dance performances. The collection, that was specially recorded for the project, is composed of professional dancers of any style or technique improvising within a precise context : 2 minutes, defined space, exact lighting. We describe a dedicated use...

This paper presents a prototype tool for browsing through multimedia libraries using content-based multimedia information retrieval techniques. It is composed of several groups of components for multimedia analysis, data mining, interactive visualization, as well as connection with external hardware controllers. The musical application of this tool...

In this article we present a system for content-based browsing of a dance video database. A set of features describing dance is proposed, to quantify local gestures of the dancer as well as global stage usage. These features are used to compute similarities between recorded dance improvisations, which in turn serve to guide the visual exploration i...

Electron microscopy of a macromolecular structure can lead to three-dimensional reconstructions with resolutions that are typically in the 30–10 Å range and sometimes even beyond 10 Å. Fitting atomic models of the individual components of the macromolecular structure (e.g. those obtained by X-ray crystallo­graphy or nuclear magnetic resonance) into...

This paper reports on the LaughterCycle project, held during a three month period between April and June 2009, within the numediart research programme. In this project, we have been developing technological building blocks for an application allow- ing to record and retrieve laughs according to their similarity. After an introduction regarding the...

The aim of this project is to explore new user interfaces and data organization for database navigation within the MediaCycle soft- ware. Both issues are deeply interconnected since the database display depends on the internal data organization and vice-versa, the data organization has to be chosen to allow useful and mean- ingful displays as well...

In order to preserve their genome integrity, organisms have developed elaborate tactics for genome protection and repair. The Deinococcus radiodurans bacteria famous for their extraordinary tolerance toward high doses of radiations or long period of desiccation, possess some specific genes with unknown function which are related to their survival i...

Rotaviruses are prototypical double-stranded RNA viruses whose triple-layered icosahedral capsid constitutes transcriptional machinery activated by the release of the external layer. To understand the molecular basis of this activation, we studied the structural interplay between the three capsid layers by electron cryo-microscopy and digital image...

The Augmented Virtual Studio (AVS) project aims at acquiring the tools of video analysis and visualization needed to achieve advanced interfaces or interaction with virtual avatars and virtual worlds. Those techniques consist in data visualization, object seg-mentation, tracking and identification of blobs but also sketch recognition and more gener...

The metric of the SO(3) group of rotations can be used to define the angular resolution of a function of rotations. The resolution is related to the degree of the highest representation present in the expansion of the function in terms of Wigner functions. The peculiar non-Euclidean metric of the rotation domain, however, implies that the terms whi...

Many bioactive peptides require amidation of their carboxy terminus to exhibit full biological activity. Peptidylglycine alpha-hydroxylating monooxygenase (PHM; EC 1.14.17.3), the enzyme that catalyzes the first of the two steps of this reaction, is composed of two domains, each of which binds one copper atom (CuH and CuM). The CuM site includes Me...

The ability to compute intra- and inter-molecular interactions provides the opportunity to gain a deeper understanding of previously intractable problems in biochemistry and biophysics. This review presents three examples in which molecular dynamics calculations were used to gain insight into the atomic detail underlying important experimental obse...

Molecular associations in solution are opposed by the loss of entropy (DeltaS) that results from the restriction of motion of each component in the complex. Theoretical estimates of DeltaS are essential for rationalizing binding affinities, as well as for calculating entropic contribution to enzyme catalysis. Recently a statistical-mechanical frame...

Blocking HIV-1 viral entry into the host cell offers a promising new strategy for interfering with the HIV-1 life cycle. A major target of inhibitor design is to prevent binding of fusogenic gp41 C-peptides to the trimeric coiled coil of fusion-active N-peptides. Here, we map the hydrophobic character of the binding surface of the IQN17 peptide, a...

This paper presents a numediart project in collaboration with two artistic projects: Méta-crâne by Thomas Israel (15) and HUM by François Zajéga (30). The scope of this project was to offer technological forecasting and development consultancy to these two highly-interactive installations that both share a common goal of blending 'behavioral" recog...

This project aims at providing tools to develop interactions with a collection of prerecorded videos and reactive sounds. These tools allow the detection of similar frames among a collection of video, the detection of visitor's behavior in a given setup, and the use of the latter information to trigger or govern video navigation and audio interacti...

Vita. U.M.I. no. 3172697. Thesis (Ph. D.)--Johns Hopkins University, 2005. Includes bibliographical references (l. 126-146) and index. Microfilm.