Search

Facial expression perception in humans inherently relies on prior knowledge and contextual cues, contributing to efficient and flexible processing. For instance, multi-modal emotional context (such as voice color, affective text, body pose, etc.) can prompt people to perceive emotional expressions in objectively neutral faces. Drawing inspiration from this, we introduce a novel approach for facial expression classification that goes beyond simple classification tasks. Our model accurately classifies a perceived face and synthesizes the corresponding mental representation perceived by a human when observing a face in context. With this, our model offers visual insights into its internal decision-making process. We achieve this by learning two independent representations of content and context using a VAE-GAN architecture. Subsequently, we propose a novel attention mechanism for context-dependent feature adaptation. The adapted representation is used for classification and to generate a context-augmented expression. We evaluate synthesized expressions in a human study, showing that our model effectively produces approximations of human mental representations. We achieve State-of-the-Art classification accuracies of 81.01% on the RAVDESS dataset and 79.34% on the MEAD dataset. We make our code publicly available., Comment: GCPR 2024

Human communication is multi-modal; e.g., face-to-face interaction involves auditory signals (speech) and visual signals (face movements and hand gestures). Hence, it is essential to exploit multiple modalities when designing machine learning-based facial expression recognition systems. In addition, given the ever-growing quantities of video data that capture human facial expressions, such systems should utilize raw unlabeled videos without requiring expensive annotations. Therefore, in this work, we employ a multitask multi-modal self-supervised learning method for facial expression recognition from in-the-wild video data. Our model combines three self-supervised objective functions: First, a multi-modal contrastive loss, that pulls diverse data modalities of the same video together in the representation space. Second, a multi-modal clustering loss that preserves the semantic structure of input data in the representation space. Finally, a multi-modal data reconstruction loss. We conduct a comprehensive study on this multimodal multi-task self-supervised learning method on three facial expression recognition benchmarks. To that end, we examine the performance of learning through different combinations of self-supervised tasks on the facial expression recognition downstream task. Our model ConCluGen outperforms several multi-modal self-supervised and fully supervised baselines on the CMU-MOSEI dataset. Our results generally show that multi-modal self-supervision tasks offer large performance gains for challenging tasks such as facial expression recognition, while also reducing the amount of manual annotations required. We release our pre-trained models as well as source code publicly, Comment: The paper will appear in the CVPR 2024 workshops proceedings

The currently deployed contact-tracing mobile apps have failed as an efficient solution in the context of the COVID-19 pandemic. None of them has managed to attract the number of active users required to achieve an efficient operation. This urges the research community to re-open the debate and explore new avenues that lead to efficient contact-tracing solutions. This paper contributes to this debate with an alternative contact-tracing solution that leverages already available geolocation information owned by BigTech companies with very large penetration rates in most countries adopting contact-tracing mobile apps. Moreover, our solution provides sufficient privacy guarantees to protect the identity of infected users as well as precluding Health Authorities from obtaining the contact graph from individuals., Comment: 7 pages, 1 figure, 1 table; added acknowledgments

Future fully interconnected virtual reality (VR) systems and the Tactile Internet diminish the boundary between virtual (online) and real (offline) worlds, while extending the digital and physical capabilities of humans via edge computing and teleoperated robots, respectively. In this paper, we focus on the Internet of No Things as an extension of immersive VR from virtual to real environments, where human-intended Internet services - either digital or physical - appear when needed and disappear when not needed. We first introduce the concept of integrated online-to-offline (O2O) communications, which treats online and offline channels as complementary to bridge the virtual and physical worlds and provide O2O multichannel experiences. We then elaborate on the emerging extended reality (XR), which brings the different forms of virtual/augmented/mixed reality together to realize the entire reality-virtuality continuum and, more importantly, supports human-machine interaction as envisioned by the Tactile Internet, while posing challenges to conventional handhelds, e.g., smartphones. Building on the so-called invisible-to-visible (I2V) technology concept, we present our extrasensory perception network (ESPN) and investigate how O2O communications and XR can be combined for the nonlocal extension of human "sixth-sense" experiences in space and time. We conclude by putting our ideas in perspective of the 6G vision., Comment: 8 pages, 4 pages, 2 tables

Visible light communication (VLC) uses huge license-free spectral bandwidth of visible light for high-speed wireless communication. Since each VLC access point covers a small area, handovers of mobile users are inevitable. In order to deal with these handovers, developing fast and effective resource allocation algorithms is a challenge. This paper introduces the problem of mobility-aware optimization of resources in VLC networks using the T-step look-ahead policy and employing the notion of handover efficiency. It is shown that the handover efficiency can correlate the overall performance of the network with future actions based on the mobility of users. Due to the stationary nature of indoor optical wireless channels, future channel state information (CSI) can be predicted by anticipating the future locations of users. A resource allocation algorithm is thus proposed, which uses CSI to dynamically allocate network resources to users. To solve this mathematically intractable problem, a novel relaxation method is proposed, which proves to be a useful tool to develop efficient algorithms for network optimization problems with a proportional fairness utility function. The resulting algorithm is extremely faster than the previous method and awareness of mobility enhances the overall performance of the network in terms of rate and fairness utility function., Comment: 13 pages and 9 figures, Journal of Lightwave Technology, 2018

In this paper, we study the cell planning problem for a two-tier cellular network containing two types of base stations (BSs)-- i.e. with fiber backhaul, referred to as wired BSs (W-BSs), and BSs with wireless backhaul, referred to as unwired-BSs (U-BSs). In-band full-duplex wireless communications is used to connect U-BSs and W-BSs. We propose an algorithm to determine the minimum number of W-BSs and U-BSs to satisfy given cell and capacity coverage constraints. Furthermore, we apply our proposed non-dominated sorting genetic algorithm II (NSGA-II) to solve both cell planning and joint cell and backhaul planning problem to minimize the cost of planning, while maximizing the coverage simultaneously. Additionally, the considered cell planning program is developed into an optimization by including the problem of minimizing the cost of fiber backhaul deployment. In order to analyze the performance of the proposed algorithm, we study three different deployment scenarios based on different spatial distributions of users and coverage areas. The results show the superiority of our proposed NSGA-II algorithm for both cell planning and joint cell and backhaul planning to other well-known optimization algorithms. The results also reveal that there is a trade-off between cell deployment costs and SINR/rate coverage, and W-BSs are placed in congested areas to consume less resources for wireless backhauls. Similarly, a trade-off between cell and fiber deployment costs and SINR/rate coverage is observed in planning. We show that for realistic scenarios desirable solutions can be selected from the Pareto front of the introduced multi-objective problem based on given cellular operator policies.

In this article, a multi-simulation model is proposed to measure the performance of all Smart Grid perspectives as defined in the IEEE P2030 standard. As a preliminary implementation, a novel information technology (IT) and communication multi-simulator is developed following an High Level Architecture (HLA). To illustrate the usefulness of such a multi-simulator, a case study of a distribution network operation application is presented using real-world topology configurations with realistic communication traffic based on IEC 61850. The multi-simulator allows to quantify, in terms of communication delay and system reliability, the impacts of aggregating all traffic on a low-capacity wireless link based on Digital Mobile Radio (DMR) when a Long Term Evolution (LTE) network failure occurs. The case study illustrates that such a multi-simulator can be used to experiment new smart grid mechanisms and verify their impacts on all smart grid perspectives in an automated manner. Even more importantly, multi-simulation can prevent problems before modifying/upgrading a smart grid and thus potentially reduce utility costs., Comment: 20 pages, 11 figures, 1 table

Building on the Ethernet Passive Optical Network (EPON) and Gigabit PON (GPON) standards, Next-Generation (NG) PONs (i) provide increased data rates, split ratios, wavelengths counts, and fiber lengths, as well as (ii) allow for all-optical integration of access and metro networks. In this paper we provide a comprehensive probabilistic analysis of the capacity (maximum mean packet throughput) and packet delay of subnetworks that can be used to form NG-PONs. Our analysis can cover a wide range of NG-PONs through taking the minimum capacity of the subnetworks making up the NG-PON and weighing the packet delays of the subnetworks. Our numerical and simulation results indicate that our analysis quite accurately characterizes the throughput-delay performance of EPON/GPON tree networks, including networks upgraded with higher data rates and wavelength counts. Our analysis also characterizes the trade-offs and bottlenecks when integrating EPON/GPON tree networks across a metro area with a ring, a Passive Star Coupler (PSC), or an Arrayed Waveguide Grating (AWG) for uniform and non-uniform traffic. To the best of our knowledge, the presented analysis is the first to consider multiple PONs interconnected via a metro network., Comment: Technical Report, School of Electrical, Computer, and Energy Eng., Arizona State University, Tempe, AZ

Current Gigabit-class passive optical networks (PONs) evolve into next-generation PONs, whereby high-speed 10+ Gb/s time division multiplexing (TDM) and long-reach wavelength-broadcasting/routing wavelength division multiplexing (WDM) PONs are promising near-term candidates. On the other hand, next-generation wireless local area networks (WLANs) based on frame aggregation techniques will leverage physical layer enhancements, giving rise to Gigabit-class very high throughput (VHT) WLANs. In this paper, we develop an analytical framework for evaluating the capacity and delay performance of a wide range of routing algorithms in converged fiber-wireless (FiWi) broadband access networks based on different next-generation PONs and a Gigabit-class multi-radio multi-channel WLAN-mesh front-end. Our framework is very flexible and incorporates arbitrary frame size distributions, traffic matrices, optical/wireless propagation delays, data rates, and fiber faults. We verify the accuracy of our probabilistic analysis by means of simulation for the wireless and wireless-optical-wireless operation modes of various FiWi network architectures under peer-to-peer, upstream, uniform, and nonuniform traffic scenarios. The results indicate that our proposed optimized FiWi routing algorithm (OFRA) outperforms minimum (wireless) hop and delay routing in terms of throughput for balanced and unbalanced traffic loads, at the expense of a slightly increased mean delay at small to medium traffic loads., Comment: Technical Report, School of Electrical, Computer, and Energy Eng. Arizona State University, Tempe