Detecting individual role using features extracted from speaker diarization results

In the field of automatic audiovisual content-based indexing and structuring, finding events like interviews, debates, reports, or live commentaries requires to bridge the gap between low-level feature extraction and such high-level event detection. In our work, we consider that detecting speaker roles like Anchor, Journalist and Other is a first step to enrich interaction sequences between speakers. Our work relies on the assumption of the existence of clues about speaker roles in temporal, prosodic and basic signal features extracted from audio files and from speaker segmentations. Each speaker is therefore represented by a 36-feature vector. Contrarily to most of the state-of-the-art propositions we do not use the structure of the document to recognize the roles of the interveners. We investigate the influence of two dimensionality reduction techniques (Principal Component Analysis and Linear Discriminant Analysis) and different classification methods (Gaussian Mixture Models, K-nearest neighbours and Support Vectors Machines). Experiments are done on the 13-h corpus of the ESTER2 evaluation campaign. The best result reaches about 82% of well recognized roles. This corresponds to more than 89% of speech duration correctly labelled.     

Inexact graph matching by means of estimation of distribution algorithms

Estimation of distribution algorithms (EDAs) are a quite recent topic in optimization techniques. They combine two technical disciplines of soft computing methodologies: probabilistic reasoning and evolutionary computing. Several algorithms and approaches have already been proposed by different authors, but up to now there are very few papers showing their potential and comparing them to other evolutionary computational methods and algorithms such as genetic algorithms (GAs). This paper focuses on the problem of inexact graph matching which is NP-hard and requires techniques to find an approximate acceptable solution. This problem arises when a nonbijective correspondence is searched between two graphs. A typical instance of this problem corresponds to the case where graphs are used for structural pattern recognition in images. EDA algorithms are well suited for this type of problems.

This paper proposes to use EDA algorithms as a new approach for inexact graph matching. Also, two adaptations of the EDA approach to problems with constraints are described as two techniques to control the generation of individuals, and the performance of EDAs for inexact graph matching is compared with the one of GAs.     

Facial emotional classification: from a discrete perspective to a continuous emotional space

User emotion detection is a very useful input to develop affective computing strategies in modern human computer interaction. In this paper, an effective system for facial emotional classification is described. The main distinguishing feature of our work is that the system does not simply provide a classification in terms of a set of discrete emotional labels, but that it operates in a continuous 2D emotional space enabling a wide range of intermediary emotional states to be obtained. As output, an expressional face is represented as a point in a 2D space characterized by evaluation and activation factors. The classification method is based on a novel combination of five classifiers and takes into consideration human assessment for the evaluation of the results. The system has been tested with an extensive universal database so that it is capable of analyzing any subject, male or female of any age and ethnicity. The results are very encouraging and show that our classification strategy is consistent with human brain emotional classification mechanisms.     

SmartTools: A Generator of Interactive Environments Tools

SmartTools is a development environment generator that provides a structure editor and semantic tools as main features. The well-known visitor pattern technique is commonly used for designing semantic analysis, it has been automated and extended. SmartTools is easy to use thanks to its graphical user interface designed with the Java Swing APIs. It is built with an open architecture convinient for a partial or total integration of SmartTools in other environments. It makes the addition of new software components in SmartTools easy. As a result of the modular architecture, we built a distributed instance of SmartTools which required minimal effort. Being open to the XML technologies offers all the features of SmartTools to any language defined with those technologies. But most of all, with its open architecture, SmartTools takes advantage of all the developments made around those technologies, like DOM, through the XML APIs. The fast development of SmartTools (which is a young project, one year old) validates our choices of being open and generic. The main goal of this tool is to provide help and support for designing software development environments for programming languages as well as application languages defined with XML technologies.     

On the ternary spatial relation "between".

The spatial relation "between" is a notion which is intrinsically both fuzzy and contextual, and depends, in particular, on the shape of the objects. The literature is quite poor on this and the few existing definitions do not take into account these aspects. In particular, an object B that is in a concavity of an object A1 not visible from an object A2 is considered between A1 and A2 for most definitions, which is counter intuitive. Also, none of the definitions deal with cases where one object is much more elongated than the other. Here, we propose definitions which are based on convexity, morphological operators, and separation tools, and a fuzzy notion of visibility. They correspond to the main intuitive exceptions of the relation. We distinguish between cases where objects have similar spatial extensions and cases where one object is much more extended than the other. Extensions to cases where objects, themselves, are fuzzy and to three-dimensional space are proposed as well. The original work proposed in this paper covers the main classes of situations and overcomes the limits of existing approaches, particularly concerning nonvisible concavities and extended objects. Moreover, the definitions capture the intrinsic imprecision attached to this relation. The main proposed definitions are illustrated on real data from medical images.     

Gas-liquid and liquid-solid mass transfers in two types of stationary catalytic basket laboratory reactor

Measurements of gas-liquid and liquid-solid mass transfer coefficients were performed in two types of three-phase laboratory reactor equipped with stationary catalytic basket and multiple impeller. Those reactors are called Robinson-Mahoney reactors. Local liquid-solid mass transfer coefficients were measured using naphtol particle dissolution in n-heptane at several agitation speeds. Experiments had shown that local coefficients depended on the particle locations in the basket and agitation speeds. The local coefficient values and profiles were quite similar for both studied reactors even though the reactor designs were different. Similar values of RM basket thicknesses might account for this phenomenon because the liquid-solid mass transfers could be correlated with the ones in packed beds with forced liquid circulation. Gas-liquid mass transfer coefficients were measured using standard dynamic absorption methods in air-water or hydrogen-gas oil system. As in stirred tanks, the gas-liquid mass transfers were linked to the impeller designs and characteristics (gas-inducing or classical impeller) and to the operating conditions such as agitation speeds and fluid properties. However, correlations applicable to stirred tank reactors can be used with Robinson-Mahoney reactors, the baskets behaving like additional baffles or reduction of tank diameters.     

High aspect ratio submicrometer two-dimensional structures fabricated by one-photon absorption direct laser writing

We have demonstrated a simple technique, based on a combination of a low cost one-photon elaboration method in a very low absorption regime (LOPA) and a tightly focusing optical system, to fabricate submicrometer 2D and 3D structures. A simple continuous-wave laser at 532 nm with only a few milliwatts allowed to fabricate high-aspect-ratio 2D pillars arrays in a commercial SU8 photoresist. The diameter of pillars is about 300 nm and the aspect ratio is as high as 7. This direct laser writing technique based on the LOPA approach is potentially a breakthrough: it is very simple, compact and low cost, while it allows to achieve the same results as those obtained by the two-photon absorption technique.     

Synthesis of new functionalized polymers and their use as stabilizers of Pd,Pt, and Rh nanoparticles. Preliminary catalytic studies

Polymers I ‐ X with functionalized main chain were synthesized from commercial polymers containing reactive groups, anhydride (for Gantrez), or carboxylate (for polyacrylates) moieties, to be used as stabilizers of metallic nanoparticles. ¹H HRMAS, ¹³C MAS NMR, IR, and GPC techniques were applied to characterize the new polymers. Metal transition nanoparticles stabilized with these polymers were prepared in organic solvents (Pd and Pt) and also in water/ethanol medium (Rh), under hydrogen atmosphere. TEM analysis revealed the formation of small nanoparticles, in the range ～ 3–20 nm, depending on the solvent, metal precursor, metal concentration, and polymer nature. Preliminary catalytic applications, C? C couplings using palladium and hydrogenations using platinum and rhodium systems in organic and also aqueous medium, were tested. © 2007 Wiley Periodicals, Inc. J Appl Polym Sci 2007