In this chapter, we motivate the use of densely-sampled light fields as the representation which can bring the required density of light rays for the correct recreation of 3D visual cues such as focus and continuous parallax and can serve as an intermediary between light field sensing and light field display. We consider the problem of reconstructing such a representation from few camera views and approach it in a sparsification framework. More specifically, we demonstrate that the light field is well structured in the set of so-called epipolar images and can be sparsely represented by a dictionary of directional and multi-scale atoms called shearlets. We present the corresponding regularization method, along with its main algorithm and speed-accelerating modifications. Finally, we illustrate its applicability for the cases of holographic stereograms and light field compression.

The light field and holographic displays constitute two important categories of advanced three-dimensional displays that are aimed at delivering all physiological depth cues of the human visual system, such as stereo cues, motion parallax, and focus cues, with sufficient accuracy. As human observers are the end-users of such displays, the delivered spatial information (e.g., perceptual spatial resolution) and view-related image quality factors (e.g., focus cues) are significantly dependent on the characteristics of the human visual system. Retinal image formation models enable rigorous characterization and subsequently efficient design of light field and holographic displays. In this chapter the ray-based near-eye light field and wave-based near-eye holographic displays are reviewed, and the corresponding retinal image formation models are discussed. In particular, most of the discussion is devoted to characterization of the perceptual spatial resolution and focus cues.

In the field of child-robot interaction (CRI), long-term field studies with users in authentic contexts are still rare. This paper reports the findings from a 4-month field study of robot-assisted language learning (RALL). We focus on the learning experiences of primary school pupils with a social, persuasive robot, and the experiences of the teachers of using the robot as a teaching tool. Our qualitative research approach includes interviews, observations, questionnaires and a diary as data collection methods, and affinity diagram as a data analysis method. The research involves three target groups: the pupils of a 3rd grade class (9–10 years old, n = 20), language teachers (n = 3) and the parents (n = 18). We report findings on user experience (UX), the robot’s tasks and role in the school, and the experience of the multimodal interaction with the robot. Based on the findings, we discuss several aspects concerning the design of persuasive robotics on robot-assisted learning and CRI, for example the benefits of robot-specific ways of rewarding, the value of the physical embodiment and the opportunities of the social role adopted by the learning robot.

This paper presents some less known details about the work of Yasuo Komamiya in development of the first relay computers using the theory of computing networks that is based on the former work of Oohashi Kan-ichi and Mochiori Goto at the Electrotechnical Laboratory (ETL) of Agency of Industrial Science and Technology, Tokyo, Japan. The work at ETL in the same direction was performed under guidance of Mochinori Goto.

For video applications, tiled streaming is a popular way to deliver viewport dependent 360-degree video. Unfortunately, dynamic adaptation to network bandwidth fluctuations of such video streams is still a challenge. This paper proposes a method for managing in a controlled way the graceful quality degradation in DASH-based streaming systems to deliver omnidirectional video. The method is enabled by the signaling of tile priority maps in order to reduce the impact of graceful degradation on the users’ Quality of Experience (QoE). Simulation results show that this method allows degrading the system by over 10% of minor bandwidth usage during Viewport Dependent Streaming, without sacrificing the user QoE so much compared to the case that does not make use of this technique. Furthermore, the presented method improves flexibility from the service provider’s standpoint.

The present paper illustrates that the game-based implementation of a learning task - here to train basic math skills - entails benefits with strings attached. We developed a game for learning math with its core element based on the number line estimation task. In this task, participants have to indicate the position of a target number on a number-line, which is thought to train basic numerical skills. Participants completed both the game on a mobile device and a conventional paper-pencil version of the task. They indicated to have significantly more fun using the game-based environment. However, they also made considerably higher estimation errors in the game compared to the paper-pencil version. In this case, more fun in a math-learning task was ultimately bought at the expense of lower reliability, namely lowered accuracy of estimations in the learning game. This fun-accuracy trade-off between adding elements for enjoyment and clarity of content is discussed together with the consequences for game-design.

The recently standardized millimeter wave-based 3GPP New Radio technology is expected to become an enabler for both enhanced Mobile Broadband (eMBB) and ultra-reliable low latency communication (URLLC) services specified to future 5G systems. One of the first steps in mathematical modeling of such systems is the characterization of the session resource request probability mass function (pmf) as a function of the channel conditions, cell size, application demands, user location and system parameters including modulation and coding schemes employed at the air interface. Unfortunately, this pmf cannot be expressed via elementary functions. In this paper, we develop an accurate approximation of the sought pmf. First, we show that Normal distribution provides a fairly accurate approximation to the cumulative distribution function (CDF) of the signal-to-noise ratio for communication systems operating in the millimeter frequency band, further allowing evaluating the resource request pmf via error function. We also investigate the impact of shadow fading on the resource request pmf.

The number of Unmanned Aerial Vehicle (UAV) applications is growing tremendously. The most critical ones are operations in use cases such as natural disasters, and search and rescue activities. Many of these operations are performed on water scenarios. A standalone niche covering autonomous UAV operation is thus becoming increasingly important. One of the crucial parts of mentioned operations is a technology capable to land an autonomous UAV on a moving surface vessel. This approach could not be entirely possible without precise UAV positioning. However, conventional strategies that rely on satellite localization may not always be reliable, due to scenario specifics. Therefore, the development of an independent precise landing technology is essential. In this paper, we developed the localization and landing system based on Gauss-Newton’s method, which allows to achieve the required localization accuracy.

The prospective roll out of recently standardized New Radio (NR) systems operating in millimeter wave frequency band pose unique challenges to network engineers. In this context, the support of NR-based vehicle-to-infrastructure communications is of special interest due to potentially high speeds of user equipment and semi-stochastic dynamic blockage conditions of propagation paths between UE and BR base station (BS). In this conditions even the use of advanced NR functionalities such as multiconnectivity supporting active connections to multiple BSs located nearby may not fully eliminate outages. Thus, to preserve session continuity for UEs located on vehicles a degree of LTE support might be required. In this paper, we quantify the amount of LTE support required to maintain session continuity in street deployment of NR systems supporting multiconnectivity capabilities. Particularly, we demonstrate that it is heavily affected by the traffic conditions, inter-site distance between NR BSs and the degree of multiconnectivity.

Design of attractive services for the bus travel context is important because of the aim to increase the usage of sustainable travel modes of public transportation. In bus travel, both user experience of the digital services and the broader service design context of the public transportation need to be addressed. Experience-Driven Design (EDD) can be used to take the passengers’ needs and experiences in the core of the design process. This paper presents a qualitative diary and interview study on bus travel experience with 20 passengers in two major cities in Finland. The aim of this study was to identify and communicate frequent bus passengers’ needs, experiences, values and activities as user insights to support experience-driven service design in the public transportation context. Based on the data analysis, we derived ten Travel Mindsets: Abstracted, Efficient, Enjoyer, In-control, Isolation, Observer, Off-line, Relaxed, Sensitive, and Social. To communicate the study findings on bus passengers’ travel experience, Travel Experience Personas were created. The personas include primary and secondary travel mindsets, specific needs related to bus travel, insights on mobile device usage, and target user experience (UX) goals that could enhance the personas’ travel experience. We also discuss how the personas can be used as a communicative design tool that supports EDD of novel services in the bus context.

The future 5G New Radio (NR) systems are expected to support both multicast and unicast traffic. However, these traffic types require principally different NR system parameters. Particularly, the area covered by a single antenna configuration needs to be maximized when serving multicast traffic to efficiently use system resources. This prevents the system from using the maximum allowed number of antenna elements decreasing the inter-site distance between NR base stations. In this paper, we formulate a model of NR system with multi-connectivity capability serving a mixture of unicast and multicast traffic types. We show that multi-connectivity enables a trade-off between new and ongoing session drop probabilities for both unicast and multicast traffic types. Furthermore, supporting just two simultaneously active links allows to exploit most of the gains and the value of adding additional links is negligible. We also show that the service specifics implicitly prioritize multicast sessions over unicast ones. If one needs to achieve a balance between unicast and multicast session drop probabilities, explicit prioritization mechanism is needed at NR base stations.

This paper reports results from an ongoing project that aims to develop a digital game for introducing fractions to young children. In the current study, third-graders played the Number Trace Fractions prototype in which they estimated fraction locations and compared fraction magnitudes on a number line. The intervention consisted of five 30 min playing sessions. Conceptual fraction knowledge was assessed with a paper based pre- and posttest. Additionally, after the intervention students’ fraction comparison strategies were explored with game-based comparison tasks including self-explanation prompts. The results support previous findings indicating that game-based interventions emphasizing fraction magnitudes improve students’ performance in conceptual fraction tasks. Nevertheless, the results revealed that in spite of clear improvement many students tended to use false fraction magnitude comparison strategies after the intervention. It seems that the game mechanics and the feedback that the game provided did not support conceptual change processes of students with low prior knowledge well enough and common fraction misconceptions still existed. Based on these findings we further developed the game and extended it with physical manipulatives. The aim of this extension is to help students to overcome misconceptions about fraction magnitude by physically interacting with manipulatives.

Human body parsing remains a challenging problem in natural scenes due to multi-instance and inter-part semantic confusions as well as occlusions. This paper proposes a novel approach to decomposing multiple human bodies into semantic part regions in unconstrained environments. Specifically we propose a convolutional neural network (CNN) architecture which comprises of novel semantic and contour attention mechanisms across feature hierarchy to resolve the semantic ambiguities and boundary localization issues related to semantic body parsing. We further propose to encode estimated pose as higher-level contextual information which is combined with local semantic cues in a novel graphical model in a principled manner. In this proposed model, the lower-level semantic cues can be recursively updated by propagating higher-level contextual information from estimated pose and vice versa across the graph, so as to alleviate erroneous pose information and pixel level predictions. We further propose an optimization technique to efficiently derive the solutions. Our proposed method achieves the state-of-art results on the challenging Pascal Person-Part dataset.

In 5G networks we expect femtocells, mmWave and D2D communications to take over the more typical long-range cellular architectures with pre-planned radio resources. However, as the connection length between the nodes become shorter, locating feasible, non-interfering combinations of the links becomes more and more difficult. In this paper a new approach to this problem is presented. In particular, through guided heuristic search, it is possible to locate non-interfering combinations of wireless connections in a highly effective manner. The approach enables operators to deploy centralized scheduling solutions for emerging technologies such as network-assisted WiFi-Direct and LTE Direct, and others, especially those which lack efficient medium arbitration mechanisms.

Image-to-image translation is a general name for a task where an image from one domain is converted to a corresponding image in another domain, given sufficient training data. Traditionally different approaches have been proposed depending on whether aligned image pairs or two sets of (unaligned) examples from both domains are available for training. While paired training samples might be difficult to obtain, the unpaired setup leads to a highly under-constrained problem and inferior results. In this paper, we propose a new general purpose image-to-image translation model that is able to utilize both paired and unpaired training data simultaneously. We compare our method with two strong baselines and obtain both qualitatively and quantitatively improved results. Our model outperforms the baselines also in the case of purely paired and unpaired training data. To our knowledge, this is the first work to consider such hybrid setup in image-to-image translation.

Gaming is acknowledged as a natural way of learning and established as a mainstream activity. Nevertheless, gaming performance and subjective game experience were hardly examined across adult age groups for which the game was not intended to. In contrast to serious games as specific tools against a natural, age-related decline in cognitive performance, we evaluated performance and subjective experiences of the established math learning game Semideus across three age groups from 19 to 79. Observed decline in performance in terms of processing speed were not exclusively predicted by age, but also by gaming frequency. Strongest age-related drops of processing speed were found for the middle-aged group aged 35 to 59 years. On the other hand, more knowledge-dependent performance measures like the amount of correctly solved problems remained comparably stable. According to subjective ratings, the middle-aged group experienced the game as less fluent and automatic compared to the younger and older groups. Additionally, the elderly group of participants reported fewer negative attitudes towards technology than both younger groups. We conclude that, albeit performance differences with respect to processing speed, subjective gaming experience stayed on an overall high positive level. This further encourages the use of games for learning across age.

In millimeter-wave (mmWave) networks, where faster signal attenuation is compensated by the use of highly directional antennas, the effects of high mobility may seriously harm the link quality and, hence, the overall system performance. In this paper, we study the channel access in unlicensed mmWave networks with mobile clients, with particular emphasis on initial beamforming training and beam refinement protocol as per IEEE 802.11ad/ay standard. We explicitly model beamforming procedures and corresponding overhead for directional mmWave antennas and provide a method for maximizing the average data rate over the variable length of the 802.11ad/ay beacon interval in different mobility scenarios. We illustrate the impact of the client speed and mobility patterns by examples of three variations of the discrete random walk mobility model.

Secure cloud storage is considered as one of the most important problems that both businesses and end-users take into account before moving their private data to the cloud. Lately, we have seen some interesting approaches that are based either on the promising concept of Symmetric Searchable Encryption (SSE) or on the well-studied field of Attribute-Based Encryption (ABE). Our construction, MicroSCOPE, combines both ABE and SSE to utilize the advantages of each technique. Finally, we enhance our construction with an access control mechanism by utilizing the functionality provided by SGX.

Despite the immense success of deep neural networks, their applicability is limited because they can be fooled by adversarial examples, which are generated by adding visually imperceptible and structured perturbations to the original image. Semantic segmentation is required in several visual recognition tasks, but unlike image classification, only a few studies are available for attacking semantic segmentation networks. The existing semantic segmentation adversarial attacks employ different gradient based loss functions which are defined using only the last layer of the network for gradient backpropogation. But some components of semantic segmentation networks implicitly mitigate several adversarial attacks (like multiscale analysis) due to which the existing attacks perform poorly. This provides us the motivation to introduce a new attack in this paper known as MLAttack, i.e., Multiple Layers Attack. It carefully selects several layers and use them to define a loss function for gradient based adversarial attack on semantic segmentation architectures. Experiments conducted on publicly available dataset using the state-of-the-art segmentation network architectures, demonstrate that MLAttack performs better than existing state-of-the-art semantic segmentation attacks.

Latency is an important metric of mobile applications performance. To reduce the latency, recently it was proposed to replace the standard centralized architecture of mobile applications by the mobile edge computing (MEC). Such an approach allows processing of users data closer to their location. Motivated by disaster response scenarios, in this paper we investigated the capabilities of MEC for the forwarding of first aid request as an illustrative example of P2P service discovery in an emergency situation. We proposed an analytical model of the system and executed performance evaluation using system level simulator. Our results show that the developed solution considerably reduces the request processing time. The proposed solution can be used not only for first aid but also for general purposes, e.g., searching various service providers in a certain location.

This paper studies near lossless JPEG image compression. A method of estimation of image regions masking ability (maximal level of distortions invisible for human visual system) using non-predictable energy of image regions is described. A novel method of zeroing quantized DCT coefficients of JPEG images to increase their compression ratio without introducing visible distortions is proposed. A numerical analysis of effectiveness of the proposed near lossless compression method using 300 noise free test images of TAMPERE17 database is carried out. It is shown that the proposed method provides an increase of compression ratio of JPEG images without visible distortions at about 1.35 times in average. Additionally, the proposed method results in decreasing of variability of compression ratio values for different images. It is shown that the proposed method increases minimal compression ratio for highly textured JPEG images from 1.1…1.5 times to 2 times. Carried out experiments demonstrated once again that the traditional PSNR metric does not correspond to human perception for this task.

We propose a novel approach for modeling semantic contextual relationships in videos. This graph-based model enables the learning and propagation of higher-level spatial-temporal contexts to facilitate the semantic labeling of local regions. We introduce an exemplar-based nonparametric view of contextual cues, where the inherent relationships implied by object hypotheses are encoded on a similarity graph of regions. Contextual relationships learning and propagation are performed to estimate the pairwise contexts between all pairs of unlabeled local regions. Our algorithm integrates the learned contexts into a Conditional Random Field (CRF) in the form of pairwise potentials and infers the per-region semantic labels. We evaluate our approach on the challenging YouTube-Objects dataset which shows that the proposed contextual relationship model outperforms the state-of-the-art methods.

The evolution of the Web browser has been organic, with new features introduced on a pragmatic basis rather than following a clear rational design. This evolution has resulted in a cornucopia of overlapping features and redundant choices for developing Web applications. These choices include multiple architecture and rendering models, different communication primitives and protocols, and a variety of local storage mechanisms. In this position paper we examine the underlying reasons for this historic evolution. We argue that without a sound engineering approach and some fundamental rethinking there will be a growing risk that the Web may no longer be a viable, open software platform in the long run.

Functional safety is involved in many machines, processes, and systems to mitigate risks by reducing the likelihood of the occurrence or the severity of the consequences of a hazard. The development of functional safety systems realising safety functions is typically directed by laws and standards, which set requirements on the development process and design of the system. In addition, functional safety systems often operate in a context, in which other control entities also affect the operation of the system under control. In this article, nine patterns considering the design and development functional safety systems, in terms of their architecture and co-operation with other controlling entities, are presented. The purpose of the patterns is to support the designers of functional safety systems to cope with the mentioned aspects.

Distributed control systems comprise networked computing units that monitor and control physical processes in feedback loops. Reliability of these systems is affected by dynamic and complex computing environments where connections and system configurations may change rapidly. Diverse redundancy can be effective in improving system dependability, but it is susceptible to common mode failures and development costs for design diversity are often seen as prohibitive. In this paper we present three patterns that can be used to provide light-weight form of fault tolerance to improve system dependability and resilience by providing ability to cope with unexpected events and faults. These patterns are presented together with a pattern language that shows how they relate to other fault tolerance patterns.

Due to growing throughput demands dictated by innovative media applications (e.g., 360°) video streaming, augmented and virtual reality), millimeter-wave (mmWave) wireless access is considered to be a promising technology enabler for the emerging mobile networks. One of the crucial usages for such systems is indoor public protection and disaster relief (PPDR) missions, which may greatly benefit from higher mmWave bandwidths. In this paper, we assess the performance of on-demand mmWave mesh topologies in indoor environments. The evaluation was conducted by utilizing our system-level simulation framework based on a realistic floor layout under dynamic blockage conditions, 3GPP propagation model, mobile nodes, and multi-connectivity operation. Our numerical results revealed that the use of multi-connectivity capabilities in indoor deployments allows for generally improved connectivity performance whereas the associated per-node throughput growth is marginal. The latter is due to the blockage-rich environment, which is typical for indoor layouts as it distinguishes these from outdoor cases. Furthermore, the number of simultaneously supported links at each node that is required to enhance the system performance is greater than two, thus imposing considerable control overheads.

The problem of predicting a novel view of the scene using an arbitrary number of observations is a challenging problem for computers as well as for humans. This paper introduces the Generative Adversarial Query Network (GAQN), a general learning framework for novel view synthesis that combines Generative Query Network (GQN) and Generative Adversarial Networks (GANs). The conventional GQN encodes input views into a latent representation that is used to generate a new view through a recurrent variational decoder. The proposed GAQN builds on this work by adding two novel aspects: First, we extend the current GQN architecture with an adversarial loss function for improving the visual quality and convergence speed. Second, we introduce a feature-matching loss function for stabilizing the training procedure. The experiments demonstrate that GAQN is able to produce high-quality results and faster convergence compared to the conventional approach.

In the wake of recent hardware developments, augmented, mixed, and virtual reality applications – grouped under an umbrella term of eXtended reality (XR) – are believed to have a transformative effect on customer experience. Among many XR use cases, of particular interest are crowded commuting scenarios, in which passengers are involved in in-bus/in-train entertainment, e.g., high-quality video or 3D hologram streaming and AR/VR gaming. In the case of a city bus, the number of commuting users during the busy hours may exceed forty, and, hence, could pose far higher traffic demands than the existing microwave technologies can support. Consequently, the carrier candidate for XR hardware should be sought in the millimeter-wave (mmWave) spectrum; however, the use of mmWave cellular frequencies may appear impractical due to the severe attenuation or blockage by the modern metal coating of the glass. As a result, intra-vehicle deployment of unlicensed mmWave access points becomes the most promising solution for bandwidth-hungry XR devices. In this paper, we present the calibrated results of shooting-and-bouncing ray simulation at 60 GHz for the bus interior. We analyze the delay and angular spread, estimate the parameters of the Saleh-Valenzuela channel model, and draw important practical conclusions regarding the intra-vehicle propagation at 60 GHz.

This paper presents a novel method which simultaneously learns the number of filters and network features repeatedly over multiple epochs. We propose a novel pruning loss to explicitly enforces the optimizer to focus on promising candidate filters while suppressing contributions of less relevant ones. In the meanwhile, we further propose to enforce the diversities between filters and this diversity-based regularization term improves the trade-off between model sizes and accuracies. It turns out the interplay between architecture and feature optimizations improves the final compressed models, and the proposed method is compared favorably to existing methods, in terms of both models sizes and accuracies for a wide range of applications including image classification, image compression and audio classification.

It is argued that we are witnessing a paradigmatic shift toward constructionist gaming in which students design games instead of just consuming them. However, only a limited number of studies have explored teaching of educational Game Design (GD). This paper reports a case study in which learning by designing games strategy was used to teach different viewpoints of educational GD. In order to support design activities, we proposed a CIMDELA (Content, Instruction, Mechanics, Dynamics, Engagement, Learning Analytics) framework that aims to align game design and instructional design aspects. Thirty under-graduate students participated in the gamified workshop and designed math games in teams. The activities were divided into eight rounds consisting of design decisions and game testing. The workshop activities were observed and the designed games saved. Most of the students were engaged in the design activities and particularly the approach that allowed students to test the evolving game after each round, motivated students. Observations revealed that some of the students had isolated design mindset in the beginning and they had problems to consider design decisions from game design and instructional perspectives, but team-based design activities often led to fruitful debate with co-designers and helped some students to expand their mindsets.

Cryptographic libraries often feature multiple implementations of primitives to meet both the security needs of handling private information and the performance requirements of modern services when the handled information is public. OpenSSL, the de-facto standard free and open source cryptographic library, includes mechanisms to differentiate the confidential data and its control flow, including run-time flags, designed for hardening against timing side-channels, but repeatedly accidentally mishandled in the past. To analyze and prevent these accidents, we introduce Triggerflow, a tool for tracking execution paths that, assisted by source annotations, dynamically analyzes the binary through the debugger. We validate this approach with case studies demonstrating how adopting our method in the development pipeline would have promptly detected such accidents. We further show-case the value of the tooling by presenting two novel discoveries facilitated by Triggerflow: one leak and one defect.

It is important for the inclusiveness of society that the youth actively participate in its development. Even though the means of digital participation have advanced in the past decade, there is still lack of understanding of digital participation of the youth. In this paper, we present a study on how youth aged 16–25 years perceive social and societal participation and more specifically, how youth currently participate in non-digitally and digitally. We conducted a mixed method study in a large gaming event in Finland using a questionnaire (N = 277) and face-to-face interviews (N = 25). The findings reveal that the gaming youth consider digital participation to include discussions in different social media services or web discussion forums. Creating digital content (e.g. videos) and answering surveys were also emphasized. Perceived advantages to participate digitally include the freedom regarding location and time, ease and efficiency in sharing information, and inexpensiveness. Central disadvantages include lack of commitment, anonymity, misinformation and cheating. We also found that frequently playing gamers are more likely to participate online in social activities than those who play occasionally. Youth who reported that they play strategy games were more active in civic participation than those who do not play strategy games. We discuss the implications of our findings to the design of tools for digital participation.

Utilization of Unmanned Aerial Vehicles (UAVs), also known as “drones”, has a great potential for many emerging applications, such as delivering the connectivity on-demand, providing services for public safety, or recovering after damage to the communication infrastructure. Notably, nearly any application of drones requires a stable link to the ground control center, yet this functionality is commonly added at the last moment in the design, necessitating compact antenna designs. In this work, we propose a novel electrically small antenna element based on the 3D folded loop topology, which could be easily located inside the UAV airframe, yet still delivering good isolation from the drones own noise sources. The complete manufacturing technique along with corresponding simulations/measurements are presented. Measurements and evaluations show that the proposed antenna design is an option to achieve genuinely isotropic radiation in a small size without sacrificing efficiency.

In this work, we briefly outline the core 5G air interface improvements introduced by the latest New Radio (NR) specifications, as well as elaborate on the unique features of initial access in 5G NR with a particular emphasis on millimeter-wave (mmWave) frequency range. The highly directional nature of 5G mmWave cellular systems poses a variety of fundamental differences and research problem formulations, and a holistic understanding of the key system design principles behind the 5G NR is essential. Here, we condense the relevant information collected from a wide diversity of 5G NR standardization documents (based on 3GPP Release 15) to distill the essentials of directional access in 5G mmWave cellular, which becomes the foundation for any corresponding system-level analysis.

The lack of realistic and open benchmarking datasets for pedestrian visual-inertial odometry has made it hard to pinpoint differences in published methods. Existing datasets either lack a full six degree-of-freedom ground-truth or are limited to small spaces with optical tracking systems. We take advantage of advances in pure inertial navigation, and develop a set of versatile and challenging real-world computer vision benchmark sets for visual-inertial odometry. For this purpose, we have built a test rig equipped with an iPhone, a Google Pixel Android phone, and a Google Tango device. We provide a wide range of raw sensor data that is accessible on almost any modern-day smartphone together with a high-quality ground-truth track. We also compare resulting visual-inertial tracks from Google Tango, ARCore, and Apple ARKit with two recent methods published in academic forums. The data sets cover both indoor and outdoor cases, with stairs, escalators, elevators, office environments, a shopping mall, and metro station.

One of the most well-known standards for radio frequency identification (RFID), the standard ISO 18000-6C, collects the requirements for RFID readers and tags and regulates respective communication protocols. In particular, the standard introduces the so-called Q-algorithm resolving conflicts in the channel (which occur when several RFID tags respond simultaneously). As of today, a vast amount of existing literature addresses various modifications of the Q-algorithm; however, none of them is known to significantly reduce the average identification time (i.e., the time to identify all proximate tags). In this work, we derive a lower bound for the average identification time in an RFID system. Furthermore, we demonstrate that in case of an error-free channel, the performance of the legacy Q-algorithm is reasonably close to the proposed lower bound; however, for the error-prone environment, this gap may substantially increase, thereby indicating the need for new identification algorithms.

The scarcity of resources available for commercial wireless access systems below 6 GHz coupled with constantly increasing traffic demands from the mobile users force network operators to seek additional spectrum. In addition to moving upper in the frequency band and occupying millimeter wave band with 3GPP New Radio access technology the set of solutions also includes implementing commercial LTE systems in unlicensed bands including 2.4 GHz and 5.1 GHz that are currently occupied by Wi-Fi. This technology, known as License Assisted Access (LAA), has recently received considerable attention within the 3GPP community. One of the solutions to provide fair division of air interface resources between competing technologies is to use schedule-based access, where LAA access point is in full control of shared medium and may dynamically schedule allocations to LTE and Wi-Fi traffic. The fine tuning of LAA technology requires careful understanding of various trade-offs and dependencies involved in Wi-Fi and LTE coexistence. In this paper, using the tools of the queuing theory we formulate and solve several analytical models targeting different implementation strategies of schedule-based LAA systems and traffic types of end users. We derive relevant performance characteristics including the session drop probabilities, probability that the session accepted to the system is drop before its service completion and average resource utilization of the system.

In this paper, we provide a shooting and bouncing ray (SBR) based simulation study of mmWave radio propagation at 60 GHz in a typical conference room. The room geometry, material types, and other simulation settings are verified against the results of the measurement campaign at 83 GHz in [15]. Here, we extend the evaluation scenario by randomly scattering several human-sized blockers as well as study the effects of human body blockage models. We demonstrate that multiple knife-edge diffraction (KED) models are capable of providing meaningful results while keeping the simulation duration relatively short. Moreover, we address another important scenario, where transmitters and receivers are located at the same heights and are moving according to a predefined trajectory that corresponds, for example, to device-to-device interactions or inter-user interference.

360-degree video has attracted more and more attention in recent years. However, it is a highly challenging task to transmit the high-resolution video within the limited bandwidth. In this paper, we first propose to unequally compress the cubemaps in each frame of the 360-degree video to reduce the total bitrate of the transmitted data. Specifically, a Group of Pictures (GOP) is used as a unit to alternately transmit different versions of the video. Each version consists of 3 high-quality cubemaps and 3 low-quality cubemaps. Then, the convolutional neural network (CNN) is introduced to enhance the low-quality cubemaps with the high-quality cubemaps by exploring the inter-frame similarities. It is shown in the experiment that a single CNN model can be used for various videos. The experimental results also show that the proposed method has an excellent quality enhancement compared with the benchmark in terms of PSNR, especially for videos with slow motion.

Differential operators are usually used to determine the rate of change and the direction of change of a signal modeled by a function in some appropriately selected function space. Gibbs derivatives are introduced as operators permitting differentiation of piecewise constant functions. Being initially intended for applications in Walsh dyadic analysis, they are defined as operators having Walsh functions as eigenfunctions. This feature was used in different generalizations and extensions of the concept firstly defined for functions on finite dyadic groups. In this paper, we provide a brief overview of the evolution of this concept into a particlar class of differential operators for functions on various groups.

As IoT devices become more powerful they can also become full participants of Internet architectures. For example, they can consume and provide RESTful services. However, the typical network infrastructures do not support the architecture and middleware solutions used in the cloud-based Internet. We show how systems designed with RESTful architecture can be implemented by using an IoT-specific technology called MQTT. Our example case is an application development and deployment system that can be used for remote management of IoT devices.

In device-to-device communications, the link quality indicators, such as signal-to-interference ratio (SIR) is heavily affected by mobility of users. Conventionally, the mobility model is assumed to be stationary. In this paper, we use kinetic theory to analyze evolution of probability distribution function parameters of SIR in D2D environment under non-stationary mobility of users. Particularly, we concentrate on elasticity of the SIR moments with respect to parameters of Fokker-Planck equation. The elasticity matrix for average SIR value, SIR variance and time periods, when SIR values is higher than a certain threshold are numerically constructed. Our numerical results demonstrate that the main kinetic parameter affecting SIR behavior is diffusion coefficient. The influence of the drift is approximately ten times less.

Most consumers own more than one device for accessing content from the Web. In this world Liquid Software allows users to switch the device and effortlessly continue tasks in the new device. This paper addresses on the needs and methods for transferring a user session and user information from one device to another. The identity should follow the moving application seamlessly instead of requiring repeated entering of credentials in each device. Such solution would make services that require authentication to work in a liquid fashion. The paper describes our on-going work on investigating how liquid transfer of user identity can be added to various ways of handing the user authentication.