Abstract

Abstract
In this paper, we address the irregular strip packing problem (or nesting problem) where irregular shapes have to be placed on strips representing a piece of material whose width is constant and length is virtually unlimited. We explore a constructive heuristic that relies on the use of graphical processing units to accelerate the computation of different geometrical operations. The heuristic relies on static selection processes, which assume that a sequence of pieces to be placed is defined a priori. Here, the emphasis is put on the analysis of the impact of these sequences on the global performance of the solution algorithm. Computational results on benchmark datasets are provided to support this analysis, and guide the selection of the most promising methods to generate these sequences.

Abstract
The development of Advanced Driver Assistance Systems ( ADAS) is rapidly growing. However, most of the ADAS require field test, which is expensive, unpredictable and time consuming. In this paper we propose a multiagent-based driving simulator which integrates a human factor analysis suite and enables rapid and low-cost experimentation of mobile-device ADAS. Our architecture uses a microscopic simulator and a serious-game-based driving simulator. The latter allows the user to control a vehicle and change the correspondent simulation state in the microscopic simulator. The driving simulator also connects to an Android device and sends several kinds of data, such as current GPS coordinates or transportation network data. One important feature of this architecture is its suitability to serve as an appropriate means to conduct behaviour elicitation through peer-designed agents, so as to improve modelling of various driving styles accounting for different aspects of preferences and perception abilities, as well as other performance measures related to drivers' interaction with ADAS solutions. The potentials of our approach to aid experiments in human factor analysis are still to be tested, but are undoubtedly huge and encouraging.

Abstract
This paper proposes a haptic interaction system for Virtual Reality (VR) based on a combination of tracking devices for hands and objects and a real-to-virtual mapping system for user redirection. In our solution the user receives haptic stimuli by manipulating real objects mapped to virtual objects. This solution departs from systems that rely on haptic devices (e.g., haptic gloves) as interfaces for the user to interact with objects in the Virtual Environment (VE). As such, the proposed solution makes use of direct haptics (touching) and redirection techniques to guide the user through the virtual environment. Using the mapping framework, when the user touches a virtual object in the VE, he will simultaneously be physically touching the equivalent real object. A relevant feature of the framework is the possibility to define a warped mapping between the real and virtual worlds, such that the relation between the user and the virtual space can be different from the one between the user and the real space. This is particularly useful when the application requires the emulation of large virtual spaces but the physical space available is more confined. To achieve this, both the user's hands and the objects are tracked. In the presented prototype we use a head-mounted depth sensor (i.e., Leap Motion) and a depth-sensing camera (i.e., Kinect). To assess the feasibility of this solution, a functional prototype and a room setup with core functionality were implemented. The test sessions with users evaluated the mapping accuracy, the user execution time and the awareness of the user regarding the warped space when performing tasks with redirection. The results gathered indicate that the solution can be used to provide direct haptic feedback in VR applications and for warping space perception within certain limits.

Abstract
Despite the rising number of emotional state detection methods motivated by the popularity increase in affective computing techniques in recent years, they are yet faced with subject and domain transferability issues. In this paper, we present an improved methodology for modelling individuals' emotional states in multimedia interactive environments. Our method relies on a two-layer classification process to classify Arousal and Valence based on four distinct physiological sensor inputs. The first classification layer uses several regression models to normalize each of the sensor inputs across participants and experimental conditions, while also correlating each input to either Arousal or Valence - effectively addressing the aforementioned transferability issues. The second classification layer then employs a residual sum of squares-based weighting scheme to merge the various regression outputs into one optimal Arousal/Valence classification in real-time, while maintaining a smooth prediction output. The presented method exhibited convincing accuracy ratings - 85% for Arousal and 78% for Valence -, which are only marginally worse than our previous non-real-time approach.

Abstract
Current affective response studies lack dedicated data analysis procedures and tools for automatically annotating and triangulating emotional reactions to game-related events. The development of such a tool would potentially allow for both a deeper and more objective analysis of the emotional impact of digital media stimuli on players, as well as towards the rapid implementation of this type of studies. In this paper we describe the development of such a tool that enables researchers to conduct objective a posteriori analyses, without disturbing the gameplay experience, while also automating the annotation and emotional response identification process. The tool was designed in a data-independent fashion and allows the identified responses to be exported for further analysis in third-party statistical software applications.