Category-level object recognition is a fundamental capability for the potential use of robots in the assistance of humans in useful tasks. There have been numerous vision-based object recognition systems yielding fast and accurate results in constrained environments. However, by depending on visual cues, these techniques are susceptible to object variations in size, lighting, rotation, and pose, all of which cannot be avoided in real video data. Thus, the task of object recognition still remains very challenging. My thesis work builds upon the fact that robots can observe humans interacting with the objects in their environment. We refer to the set of objects, which can be involved in the interaction as `interactionable' objects. The interaction of humans with the `interactionable' objects provides numerous nonvisual cues to the identity of objects. In this thesis, I will introduce a flexible object recognition approach called Multiple-Cue Object Recognition (MCOR) that can use multiple cues of any predefined type, whether they are cues intrinsic to the object or provided by observation of a human. In pursuit of this goal, the thesis will provide several contributions: A representation for the multiple cues including an object definition that allows for the flexible addition of these cues; Weights that reflect the various strength of association between a particular cue and a particular object using a probabilistic relational model, as well as object displacement values for localizing the information in an image; Tools for defining visual features, segmentation, tracking, and the values for the non-visual cues; Lastly, an object recognition algorithm for the incremental discrimination of potential object categories. We evaluate these contributions through a number of methods including simulation to demonstrate the learning of weights and recognition based on an analytical model, an analytical model that demonstrates the robustness of the MCOR framework, and recognition results on real video data using a number of datasets including video taken from a humanoid robot (Sony QRIO), video captured from a meeting setting, scripted scenarios from outside universities, and unscripted TV cooking data. Using the datasets, we demonstrate the basic features of the MCOR algorithm including its ability to use multiple cues of different types. We demonstrate the applicability of MCOR to an outside dataset. We show that MCOR has better recognition results over vision-only recognition systems, and show that performance only improves with the addition of more cue types.

Šiame darbe mes apžvelgėme modalumo logiką S4 bei kvantorinę modalumo logiką S4. Taip pat jų taisykles, aksiomas ir naudojamus skaičiavimus. Pateikėme kelias sekvencijų išvedimo pavyzdžių. Taip pat, apžvelgėme kai kurių, atskirų šių logikų klasių išsprendžiamumą. Taipogi šiame darbe buvo nagrinėjama labai įdomi tema – išsprendžiamumo klasių formavimas, naudojantis formulių transformavimų į klasikinę predikatų logiką. / Logic is the branch of mathematics that deals with the formal principles, methods and criteria of validity of inference, reasoning and knowledge. Logic is concerned with what is true and how we can know whether something is true. This involves the formalization of logical arguments and proofs in terms of symbols representing propositions and logical connectives. The goal of this work is to learn more about modal logic S4 and to consider some it decidable classes of formulas. It’s important, because decidable classes helps the substantiation of different formulas. In this work we will consider the formulas of modal logic without functional symbols.

This thesis investigates logical formalizations of Castelfranchi and Falcone's (C&F) theory of trust [9, 10, 11, 12]. The C&F theory of trust defines trust as an essentially mental notion, making the theory particularly well suited for formalizations in multi-modal logics of beliefs, goals, intentions, actions, and time. Three different multi-modal logical formalisms intended for multi-agent systems are compared and evaluated along two lines of inquiry. First, I propose formal definitions of key concepts of the C&F theory of trust and prove some important properties of these definitions. The proven properties are then compared to the informal characterisation of the C&F theory. Second, the logics are used to formalize a case study involving an Internet forum, and their performances in the case study constitute grounds for a comparison. The comparison indicates that an accurate modelling of time, and the interaction of time and goals in particular, is integral for formal reasoning about trust. Finally, I propose a Horn fragment of the logic of Herzig, Lorini, Hubner, and Vercouter [25]. The Horn fragment is shown to be too restrictive to accurately express the considered case study.

The convenience of Operational Modal Analysis (OMA), over conventional Experimental Modal Analysis (EMA), has seen to its increasing popularity over the last decade for the purpose of evaluating dynamic properties of structures. OMA features an advantage of requiring only output information, which is in tandem with its main drawback of lacking scaled modeshape information. While correctly scaled modeshapes can be assumed under a restrictive assumption of spectrally white inputs, in reality, input spectra are at best broadband in nature. In this thesis, an OMA method for Multiple-Input Multiple-Output (MIMO) applications in mechanical structures is developed. The aim is to separate MIMO responses into a collection of Single-Input Single-Output (SISO) processes (matrix FRF) using cepstral-based methods, under less restrictive and hence more realistic coloured broadband excitation. Existing cepstral curve-fitting techniques can be subsequently applied to give regenerated FRFs with correct relative scaling. This cepstral-based method is based on the matrix Mean Differential Cepstrum (MDC) and operates in the frequency domain. Application of the matrix MDC onto MIMO responses leads to a matrix differential equation which together with the use of finite differences, directly solves or identifies the matrix FRF in a propagative manner. An alternative approach based on whitened MIMO responses can be similarly formulated for the indirect solution of the matrix FRF. Both the direct and indirect approaches can be modified with a Taylor series approximation to give a total of four propagative solution sequences. The method is developed using relatively simple simulated and experimental systems, involving both impulsive and burst random excitations. Detailed analysis of the results is performed using more complicated Single-Input Multiple-Output (SIMO) and MIMO systems, involving both driving and non-driving point measurements. The use of the matrix MDC method together with existing cepstral curve-fitting technique to give correct relative scaling is demonstrated on a simulated MIMO system with coloured inputs. Accurate representation of the actual FRFs is achieved by the matrix MDC technique for SIMO set-ups. In MIMO scenarios, excellent identification was obtained for the case of simulated impulsive input while the experimental and burst random input cases were less favourable. The results show that the matrix MDC technique works in MIMO scenarios, but possible noise-related issues need to be addressed in both experimental and burst random input cases for a more satisfactory identification outcome.

Thesis (M.S.)--Rochester Institute of Technology, 1993. / Includes bibliographical references (leaves 80-87).

Dissertation (Ph. D.)--University of Akron, Dept. of Civil Engineering, 2009. / "August, 2009." Title from electronic dissertation title page (viewed 9/16/2009) Advisor, Craig C. Menzemer; Committee members, Wieslaw Binienda, Robert Liang, D. Dane Quinn, Kevin L. Kreider; Department Chair, Wieslaw Binienda; Dean of the College, George K. Haritos; Dean of the Graduate School, George R. Newkome. Includes bibliographical references.

Thesis (Ph. D.)--Virginia Polytechnic Institute and State University, 1993. / Vita. Abstract. Includes bibliographical references (leaves 135-138). Also available via the Internet.

In this paper we describe a technique of classifier combination used in a human identification system. The system integrates all available features from multi-modal sources within a Bayesian framework. The framework allows representinga class of popular classifier combination rules and methods within a single formalism. It relies on a “per-class” measure of confidence derived from performance of each classifier on training data that is shown to improve performance on a synthetic data set. The method is especially relevant in autonomous surveillance setting where varying time scales and missing features are a common occurrence. We show an application of this technique to the real-world surveillance database of video and audio recordings of people collected over several weeks in the office setting.

Thesis (M.S.)--Rochester Institute of Technology, 1993. / Typescript. Includes bibliographical references (leaves 100-102).

Populärmusik svämmar över av harmoniska förlopp hämtade ur både folk- och konstmusik, vilket har föranlett att inget samförstånd finns kring huruvida sådan musik bör analyseras. Det har emellertid gjorts analyser av angloamerikansk populärmusik utifrån neo-Riemanniansk teori, där musikanalytikerna menat att detta förfarande haft god förklaringskraft. Ambitionen med denna studie var att bidra till neo-Riemanniansk teoriutveckling. Detta genom att analysera ackordprogressioner från fyra svenska populärmusikaliska verk, där vedertagna analysverktyg, som exempelvis funktionsanalys, inte antogs vara fruktbara. Analyserna av ackordprogressionerna gjordes med utgångspunkt i neo-Riemanniansk teori, därefter tonalt (funktionsanalys) och modalt. Detta trefaldiga tillvägagångssätt ansågs vara relevant för att visa på olika analysmetoders fruktbarhet. Resultatet visade att neo-Riemanniansk teori enbart kan förklara förhållandet mellan två konsonanta treklanger, dock alla ackordförhållanden i denna analys, men inte ett musikstycke i sin helhet. Vid analys av låtar med septimackord hade ett neo-Riemannianskt förfarande inte god förklaringskraft, där i stället funktionsanalys kunde förklara ackordprogressioner bättre. Dessa två analysverktyg kompletterade varandra, medan musikexemplen var svåra att förklara modalt.