In this paper we introduce EST-GAN, an approach to improve the realism of frames from unsophisticated game scenes. For this purpose, several Generative Adversarial Networks (GANs) structures are applied, which are extended to handle intermediate game render passes generated by conventional rendering pipelines. We present an image-to-image translation method that transforms simple low-poly game scenes into the style of an elaborately produced video game. Through our experiments, we show that the involvement of G-buffer information has a significant impact on the results of these translations and the usage of these leads to a reduction in artifacts.

Studies have shown that gamification increases motivation and user experience when it comes to a certain behavior or completing a process. Gamification is often deeply associated with naïve animations and stylized text. This paper addresses the effect that visual representation has on the motivation of a subject by measuring their motivation after completing a mundane process, with entertaining gamification elements as well as gamification elements presented in plain text. For the purposes of this study a within subject design was used to gather data. Participants completed a mundane task three times, once without any gamification elements, once with pragmatic feedback and once with entertaining feedback. After completion the participants filled out the same Likert scale survey. The results were evaluated using the Wilcoxon Signed-Ranks test method, indicated that there was not a significant difference in user motivation between the visually stylized and plain text feedback. If conducted on a larger scale, this discovery could lead to a reduction time and cost for gamification development.

School students are increasingly distracted by modern technology, leading to decreasing skills in reading and writing. This paper presents the E-learning-tool "Story Templates", which enables teachers to include stories into the daily homework assignments. The basic idea is to insert learning content meaningful into exciting stories, coming from a large database of stories. Each story has its predefined breaks, where different templates or rather tasks can be inserted, for example an arithmetic problem. Teachers can choose between various ways of grading the homework results, including multiple choice, text based etc. The tool also contains an app that students can use on their smartphones or tablets to read the stories and provide answers. The goal of this work is to make learning more fun again, by using modern technology.

In extensive trials we asked participants with disabilities to play a set of computer games, demanding a variety of cognitive and physical skills. In particular we presented a Tetris variant we developed ourselves, which follows various includification guidelines. Participants were recorded on video in order to identify strengths and weaknesses in coping with the presented game challenges. In the videos we recorded the game screen, a frontal video of the participants showing emotional responses, and eye gaze. Post processing tasks are used to analyze and retrieve data from the videos. This helps to build up a video database for later analysis and answer questions raised after the recordings were taken.

INTERACCT is a multidisciplinary research project, where a communication tool in form of a serious video game was implemented. Target group are young and adolescent patients which suffered leukaemia or similar diseases that indicate a stem cell transplantation as therapy. After the transplantation, patients are required to stay at home for 12–24 months and report their health status to their after-care medical team. To ease this reporting process, INTERACCT was conceptualized and a prototype was implemented. Through smartphones, tablets or desktop PCs the patients report their data to the medical team, which accesses this data through a newly developed web-system. The system is highly adaptable and customizable, and can be further extended using state-of-the-art technologies like Microsoft KINECT. This paper describes background information, the concept and finally the implementation of the prototype system.

Smartphone-based applications are increasingly being used to combat unhealthy habits resulting in obesity or reduced physical mobility. In turn, gamification has been identified as a useful tool to motivate individuals towards behavioural change. However, both health professionals and game developers are currently working on island solutions addressing individual problems due to the lack of efficient frameworks enabling cooperation in this domain. This has resulted in a large number of mHealth apps, each using different motivational techniques, metrics and health data. In this paper we propose a unified user-centred framework which is capable of running third-party developer's apps within a sandbox. Using this approach, we can mitigate often voiced privacy and safety concerns and simultaneously tailor health-interventions using content from developers as well as user preferences and environment in order to facilitate internalisation of healthy lifestyles. Apart from this, we provide indications for operationalizing our conceptual framework in future work.

Mobile cellular networks can serve as ubiquitous sensors for physical mobility. We propose a method to infer vehicle travel times on highways and to detect road congestion in real-time, based solely on anonymized signaling data collected from a mobile cellular network. Most previous studies have considered data generated from mobile devices active in calls, namely Call Detail Records (CDR), an approach that limits the number of observable devices to a small fraction of the whole population. Our approach overcomes this drawback by exploiting the whole set of signaling events generated by both idle and active devices. While idle devices contribute with a large volume of spatially coarse-grained mobility data, active devices provide finer-grained spatial accuracy for a limited subset of devices. The combined use of data from idle and active devices improves congestion detection performance in terms of coverage, accuracy, and timeliness. We apply our method to real mobile signaling data obtained from an operational network during a one-month period on a sample highway segment in the proximity of a European city, and present an extensive validation study based on ground-truth obtained from a rich set of reference datasources - road sensor data, toll data, taxi floating car data, and radio broadcast messages.

Detecting if two or multiple devices are moved together is an interesting problem for different applications. However, these devices may be aligned arbitrarily with regards to each other, and the three dimensions sampled by their respective local accelerometers can therefore not be directly compared. The typical approach is to ignore all angular components and only compare overall acceleration magnitudes --- with the obvious disadvantage of discarding potentially useful information. In this paper, we contribute a method to analytically determine relative spatial alignment of two devices based on their acceleration time series. Our method uses quaternions to compute the optimal rotation with regards to minimizing the mean squared error. The implication is that the reference system of one device can be (locally and independently) aligned with the other, and thus that all three dimensions can consequently be compared for more accurate classification. Based on real-world experimental data from smart phones and smart watches shaken together, we demonstrate the effectiveness of our method with a magnitude squared coherence metric, for which we show an improved EER of 0.16 (when using derotation) over an EER of 0.18 (when not using derotation).

Distributed dynamic VM consolidation can be an effective strategy to improve energy efficiency in cloud environments. In general, this strategy can be decomposed into four decision-making tasks: (1) Host overloading detection, (2) VM selection, (3) Host underloading detection, and (4) VM placement. The goal is to consolidate virtual machines dynamically in a way that optimizes the energy-performance tradeoff online. In fact, this goal is achieved when each of the aforementioned decisions are made in an optimized fashion. In this paper we concentrate on the VM selection task and propose a Fuzzy Q-Learning (FQL) technique so as to make optimal decisions to select virtual machines for migration. We validate our approach with the CloudSim toolkit using real world PlanetLab workload. Experimental results show that using FQL yields far better results w.r.t. the energy-performance trade-off in cloud data centers in comparison to state of the art algorithms.

Nowadays, unmanned aerial vehicles (UAVs) have become very popular for civil as well as military purposes, ranging from small quadcopters up to plane sized drones. UAVs are usually controlled manually or autonomously, both ways often heavily relying on the use of a global positioning system such as GPS. However, especially the wide spread GPS suffers form large positioning errors of 10-30 m on average. In this paper it is shown how UAVs can improve the accuracy of their position estimates by cooperating with each other. By using ranging hardware, UAVs can create a local positioning frame of reference, and use this to reduce the average positioning error of the global positions. This is done by finding an optimal 3D-rotation between the local and global coordinates. By using simulation, we demonstrate the feasibility of our approach, as well as the influence of parameters like vehicle speed and area size on the average positioning error.

Global positioning technologies such as GPS are ubiquitously available for modern smartphones, on-board navigation computers of vehicles, sports appliances, and so on. Positioning enables new networking and application solutions, but suffers from insufficient positioning accuracy. We present a way for improving the accuracy of a global positioning system by adding a local positioning system giving relative positions of a group of mobile entities more accurate than the global system communicated among the mobile entities. The improvement is based on transforming the global coordinate system into the local coordinate system and back by translation and rotation operations. In this paper, we detail the geometric operations and present a novel derivation of how to compute the optimal rotation angle based on covariance measures. Further, we provide estimates for the overall prediction error when using the local positioning system in addition to a global one and a proof of the convergence of the approach. Two major results are the fact that error correlations in the global positioning system increase the prediction error, while error correlations in the local positioning system decrease the overall prediction error. In a case study, we present results from randomized experiments, as well as from GPS and Differential GPS (DGPS) measurements. Both sets of experiments confirm the theoretical findings.

Road traffic can be monitored by means of static sensors and derived from floating car data, i.e., reports from a sub-set of vehicles. These approaches suffer from a number of technical and economical limitations. Alternatively, we propose to leverage the mobile cellular network as a ubiquitous mobility sensor. We show how vehicle travel times and road congestion can be inferred from anonymized signaling data collected from a cellular mobile network. While other previous studies have considered data only from active devices, e.g., engaged in voice calls, our approach exploits also data from idle users resulting in an enormous gain in coverage and estimation accuracy. By validating our approach against four different traffic monitoring datasets collected on a sample highway over one month, we show that our method can detect congestions very accurately and in a timely manner.

When watching video in a mobile environment, many distractions can draw the attention of the user away from the terminal, e.g., a smartphone. The power consumed by the phone during video playback is wasted in case the user is not watching. In this paper, we present an interactive video player allowing to temporarily interrupt the playback when the user is looking away, and resume it when the user starts looking at the phone again. In order to detect user inattention, we exploit the front-facing camera embedded in most current smartphones as image source for face detections. We analyze the power consumption of the phone when running our interactive video player in controlled and realistic scenarios, and study the energy savings induced by the attention-aware playback interruption. Additionally, we study the influence of video bitrate and content on power consumption. Finally, we propose a cloud-based approach in which the image captured by the front camera is sent to a server, which performs the face detection. We show that the power savings are significant using this approach, and can compensate the consumption added by the increased wireless traffic.

Virtualized data centers where several virtual machines (VMs) are hosted per server are becoming more popular due to Cloud Computing. As a consequence of energy efficiency concerns, the exact combination of VMs running on a specific server will most likely change over time. We present experimental results how to use the energy/power consumption logs of a power monitored server as a side-channel that allows us to recognize the exact combination of VMs it currently hosts to a high degree. For classification, we use a maximum log-likelihood approach, which works well for comparably small training and test set sizes. We also show to which degree a specific VM can be recognized, regardless of other VMs currently running on the same server, and show false negative/positive rates. To cross-validate our results, we have used a Kolmogorov-Smirnov test, resulting in comparable quality of recognition within shorter time. In order to clarify whether our approach is generalizable and yields reproducible results, we have set up a second experimental infrastructure in Lyon, using a different hardware platform and power measurement device. We have obtained similar results and have experimented with different CPU frequency scaling governors, yielding comparable quality of recognition. As a result, energy consumption data of servers must be protected carefully, as it is potentially valuable information for an attacker trying to track down a VM to mount further attack steps.

Today, mobile communication devices allow users to access a wide variety of multimedia contents and services. In order to improve user experience and device usability, the design of interfaces and interaction techniques for mobile devices have focused on new modalities, other than those used for desktop computers. In this paper, we describe a novel gesture controlled video player interface for mobile devices. The results of a usability study confirm that users would definitely like to adopt the major part of the proposed features. Furthermore, the responsiveness and reliability of the interface has been studied. Measured response times have been found to be within acceptable boundaries and the number of unrecognised haptic controls is limited.

Video patching is a way for transporting true video-on-demand, that is, instantaneous without any delay, from a video server to several clients. Instead of sending a unique stream to each newly arriving client, clients share as many multicast transmissions as possible, and are serviced only those parts of the video that they have missed. We present a novel video patching scheme using hierarchies of patches. Our scheme minimizes the bandwidth needed by the video server, and may result in the fact that clients receive several streams in parallel. We show analytically that for Poisson arrival our algorithm achieves the optimal possible server bandwidth for all schemes where clients share multicast transmissions. We also show, how our approach can be combined with batching. This combination requires less server bandwidth than all fixed start point periodic broadcast algorithms.

A novel management layer architecture for mobile grid environments and ad-hoc networks is presented. We address a key challenge in dynamic mobile environments with a high degree of fluctuation in the hardware infrastructure: how to store information persistently and at the same time efficiently. Most currently available approaches are based on replication strategies. These strategies achieve persistency even in highly dynamic environments, but have to pay a high price in terms of redundancy. Our proposed architecture is based on the adaptation of fault-tolerant data distribution strategies utilizing erasure codes, such as RAID or Reed-Solomon, and thus combines persistency with high storage efficiency.

The simulation of computer systems requires representative, reliable workload models. When simulating computer systems or components that are influenced by user behavior, this very behavior has to be modeled by using mathematical means. In this article, we propose a user behavior model framework that is constructed in a top-down manner consisting of various layers. The layers offer services to the next higher layer and require services from the layers below. The framework is meant to enable the modeler to plug in his own models at the layers of his choice, thus choosing the right balance between the simulation complexity and creating representative results. We give a description of the layered framework and the corresponding methodology. The approach is then demonstrated on modeling HTTP traffic, to be used in network traffic simulation.