The paper presents a new global optimization method based on simulated annealing (SA) and differential evolution (DE). The algorithm also includes a local search step that speeds up the convergence to a minimum of the function subject to optimization. The algorithm is designed as an asynchronous parallel system that makes it possible to evaluate multiple candidate solutions in parallel. The efficiency of the algorithm was confirmed with a comparison with SA and DE on test functions. Processing speedup using 1, 2, 4, and 8 processors was demonstrated. The proposed algorithm was applied to several integrated circuit sizing problems where it found near optimal circuits.
COBISS.SI-ID: 8151124
Modern information systems make it increasingly easy to gain more insight into the public interest, which is becoming more and more important in diverse public and corporate activities and processes. The disadvantage of existing research that focuses on mining the information from social networks and online communities is that it doesn’t uniformly represent all population groups and that the content can be subjected to self-censoring or curation. In this paper we propose and describe a framework and a method for estimating public interest from the implicit negative feedback collected from the IPTV audience. Our research focuses primarily on the channel change events and their match with the content information obtained from closed captions. The presented framework is based on concept modeling, viewership profiling, and combines the implicit viewer reactions (channel changes) into an interest score. The proposed framework addresses both above mentioned disadvantages or concerns. It is able to cover a much broader population, and it can detect even minor variations in user behavior. We demonstrate our approach on a large pseudonymized real-world IPTV dataset provided by an ISP, and show how the results correlate with different trending topics and with parallel classical long-term population surveys.
COBISS.SI-ID: 11734356
This paper presents a wearable training system designed to facilitate the learning process of proper movement patterns in sports training. The system implements a gesture user interface and real-time biofeedback. To demonstrate the concept of the proposed real-time biofeedback training system, an application for golf swing training is developed. The application implements the system using smartphone motion sensors and audio biofeedback and aids golfers in correcting unwanted head movements during a golf swing. The application is driven by a gesture user interface. During the golf swing, the application provides users with real-time audio feedback that signals head movement errors. The field test results show that the developed application can be used as an efficient tool in golf swing training.
COBISS.SI-ID: 11111764
This paper describes a user study on the interaction with an in-vehicle information system (IVIS). The motivation for conducting this research was to investigate the subjectively and objectively measured impact of using a single- or multi-modal IVIS while driving. A hierarchical, list-based menu was presented using a windshield projection (head-up display), auditory display and a combination of both interfaces. The users were asked to navigate a vehicle in a driving simulator and simultaneously perform a set of tasks of varying complexity. The experiment showed that the interaction with visual and audiovisual head-up displays is faster and more efficient than with the audio-only display. All the interfaces had a similar impact on the overall driving performance. There was no significant difference between the visual only and audio-visual displays in terms of their efficiency and safety; however, the majority of test subjects clearly preferred to use the multi-modal interface while driving.
COBISS.SI-ID: 10729812
Smartphones are currently the most pervasive wearable devices. One particular use of smartphone inertial sensors is motion tracking in various mobile systems and applications. The objective of this study is to validate smartphone gyroscopes for angular tracking in mobile biofeedback applications. The validation method includes measurements of angular motion performed concurrently by a smartphone gyroscope and a professional optical tracking system serving as the reference. The comparison of the measurement results shows that the inaccuracies of a calibrated smartphone gyroscope for various movements are between 0.42° and 1.15°. Based on the measurement results and the general requirements of biofeedback applications, smartphone gyroscopes are sufficiently accurate for angular motion tracking in mobile biofeedback applications.
COBISS.SI-ID: 11470164
This paper describes a user study on interaction with a mobile device installed in a driving simulator. Two new auditory interfaces were proposed and their effectiveness and efficiency were compared to a standard visual interface. Both auditory interfaces consisted of spatialized auditory cues representing individual items in the hierarchical structure of the menu. The visual interface was shown on a small in-vehicle LCD screen. In all three cases, a custom-made interaction device (a scrolling wheel and two buttons) was used for controlling the interface. The experiment proved that both auditory interfaces were effective to use in a mobile environment when user’s visual channel was strongly distracted.
COBISS.SI-ID: 6450004
MC (muscle contraction) sensor is based on a novel principle whereby muscle tension is measured during muscle contractions. During the measurement, the sensor is fixed on the skin surface above the muscle, while the sensor tip applies pressure and causes an indentation of the skin and intermediate layer directly above the muscle and muscle itself. The force on the sensor tip is then measured. This force is roughly proportional to the tension of the muscle. The measurement is noninvasive and selective. The sensor is relatively small and light so that the measurements can be performed while the measured subject performs different activities. The sensor will be used in sport trainings, rehabilitation after injuries, and also in diagnostics.
COBISS.SI-ID: 8712276
The paper presents an analytic derivation of the axis and angle of the single rotation equivalent to three simultaneous rotations around orthogonal axes when the measured angular velocities or their proportions are approximately constant. Based on the resulting expressions, a vector called the simultaneous orthogonal rotations angle (SORA) is defined, with components equal to the angles of three simultaneous rotations around coordinate system axes. The orientation and magnitude of this vector are equal to the equivalent single rotation axis and angle, respectively. As long as the orientation of the actual rotation axis is constant, given the SORA, the angular orientation of a rigid body can be calculated in a single step, thus making it possible to avoid computing the iterative infinitesimal rotation approximation.
COBISS.SI-ID: 8574804
The aim of the study was to examine, whether by the human observer observed attention of the learner could be computationally modelled in real time. An experimental study on the real-time estimation of observed learners' attention given the task of touch-typing was performed. The study has shown, it is possible to model learner's attention in real time from psycho-physiological parameters with relatively high sampling frequency which is impossible to achieve with traditional assessment methods (e.g. between session self-reports). Human ratings obtained by human raters watching session video were used as ground truth. The results show, that computational models were relatively successful discriminating low versus high levels of attention within learners. The influence of individual psychophysiological and affective signals (eye gaze, pupil dilation, and valence and arousal) on the estimation of the observed attention was also examined. The results show that both affective dimensions (valence and arousal), as well as eye gaze offset significantly and most frequently influence the performance of the within-learner model.
COBISS.SI-ID: 11814484
The paper we show how the usage of affective metadata, which can be acquired using physiological sensors, improves the accuracy of a recommender system for images. We used a user modeling tecnique that exploits the user's preference towards the valence, arousal and dominance dimenisoin of emotions.
COBISS.SI-ID: 7907412