A text generation model is a machine learning model that uses neural networks, especially transformers architecture to generate contextually relevant text based on linguistic patterns learned from extensive corpora. The models are trained on a huge amount of textual data so that they can model and learn complex concepts of any language like its grammar, vocabulary, phrases, and styles.

Obstacle detection is crucial for the safety and efficiency of autonomous vehicles. For mini-vehicles such as palm-sized drones, it is a challenge to implement traditional methods like Lidar due to high costs and physical constraints. Vision-based deep learning approaches, while accurate, are too resource-intensive for the mini-vehicles. To address this issue, we introduce Flowdep, a novel optical-flow-based algorithm inspired by the low-resolution but efficient motion-detection mechanisms in insects. Flowdep combines optic flow and IMU (or positioning information) to estimate the depth of every image pixel. We also generate a variant of Flowdep using the artificial neural network (Flowdep-ANN). Our tests show that Flowdep and Flowdep-ANN are 5.8 to 114.7 times faster than the DNN networks we tested, while the accuracies of Flowdep and Flowdep-ANN are on par with these networks. We further tested Flowdep and Flowdep-ANN on a small autonomous vehicle with Raspberry Pi4 as the computing platform, and both models successfully performed real-time object detection. The present work demonstrates the potential of using optical flow as an efficient approach to estimate depth and detect obstacles in resource-constrained mini-vehicles.

Sports are data-driven: individual performances are measured using statistics and teams leverage data analytics to outperform competition. Sports media-which is created by media outlets, teams, and individuals-engage its consumers by creating narratives about the sport, teams, and players. Due to the importance of data in the sports world, data visualizations are a pillar in the sports media landscape. These data visualizations, while appearing to accurately convey data to its consumers, can be misinformative; media creators often have incentives to present specific narratives which don't always fit the data. This work contributes to an existing misinformative data visualization taxonomy. In doing so, it makes it easier to understand the techniques and design choices used to create misinformative visualizations in all domains, not only in sports media.

"Radio frequency identification" (RFID) systems are ubiquitous in today's world. In an RFID system, it is a desirable to attain mutual authentication between a reader and a tag before commencing application-level communications. This is because tags should not share secret information with unknown parties and readers need to defend against tag impersonation. Authentication protocols designed for communication between computers, however, are not appropriate for RFID systems because tags are extremely resource constrained (low energy, small memory, etc.). Thus, there have been many attempts to design secure and practical authentication protocols for RFID systems over the years since RFID systems became prevalent. This survey summarizes and compares these protocols.

This article examines the effect of multipath channels on the performance of distributed arrays that employ time reversal. A model of the signal received from a distributed array is formulated, and a statistical analysis of the variation in signal power in the presence of phase noise and multipath is given. We present the impact these nonidealities have on received signal power, and we analyze the received power for three specific cases: continuous waveform, impulse waveform, and modulated rectangular pulse waveform in the presence of standard channel models. It is shown that for larger arrays in multipath channels, the change in power between coherent and incoherent states converges to the line-of-sight channel. It is further shown that in a line-of-sight channel time-reversal completely cancels unknown channel delays resulting in coherent signals from all nodes in a distributed array, while in a multipath channel only the main diagonal round-trip paths are coherent. Nevertheless, this additional benefit improves signal coherence in complex channels and can aide in distributed array synchronization using two-way time transfer.

Hodge conjecture is turned out to be false in extension of graph theory based on its algebraic attribute.

The evolution of 3D generative modeling has been notably propelled by the adoption of 2D diffusion models. Despite this progress, the cumbersome optimization process per se presents a critical hurdle to efficiency. In this paper, we introduce Hash3D, a universal acceleration for 3D generation without model training. Central to Hash3D is the insight that feature-map redundancy is prevalent in images rendered from camera positions and diffusion time-steps in close proximity. By effectively hashing and reusing these feature maps across neighboring timesteps and camera angles, Hash3D substantially prevents redundant calculations, thus accelerating the diffusion model's inference in 3D generation tasks. We achieve this through an adaptive grid-based hashing. Surprisingly, this feature-sharing mechanism not only speed up the generation but also enhances the smoothness and view consistency of the synthesized 3D objects. Our experiments covering 5 textto-3D and 3 image-to-3D models, demonstrate Hash3D's versatility to speed up optimization, enhancing efficiency by 1.3 ∼ 4×. Additionally, Hash3D's integration with 3D Gaussian splatting largely speeds up 3D model creation, reducing text-to-3D processing to about 10 minutes and image-to-3D conversion to roughly 30 seconds. The code is provided in https://github.com/Adamdad/hash3D.

Physically Unclonable Function (PUF) is an emerging hardware security primitive that provides a promising solution for lightweight security. PUFs can be used to generate a secret key that depends on the random manufacturing process variation of the device for lightweight authentication and device identification. This work proposes an optimized version of the Configurable Ring Oscillator (CRO) PUF that aims to reduce power consumption and area overhead. The proposed design eliminates the duplication of ROs, reduces the switching activity, and introduces the inter-stage delay as an additional source of randomness. The proposed PUF has been implemented in 22nm FDSOI technology using the Synopsys tools. A comprehensive security analysis has been acquired utilizing Challenge-Response Pairs collected from 8 chips. Results show an average of 49.42%, 38.25%, 9.95%, and 45.5% for uniformity, diffuseness, reliability, and uniqueness, respectively. Compared with the state-of-the-art, the proposed design achieves an area and power reduction of 75% and 65.1%, respectively. With the proposed PUF delivering 10 32 CRPs, it is classified as a strong PUF. Additionally, the proposed design passes NIST tests and achieves an average prediction accuracy of 67.1% of machine learning modeling.

SORA is a text-to-video (T2V) model that can generate realistic video content based on simple user prompts. Upon its release to the general public, SORA has the potential to revolutionize numerous industries while also presenting significant challenges. This research aims to provide a comprehensive understanding of SORA’s opportunities, challenges, and implications. It explores its potential applications in filmmaking, education, gaming, advertising, accessibility, healthcare, and social media content creation. Additionally, it investigates potential challenges such as misinformation, erosion of privacy, bias, regulatory complexities, and technological dependence. Moreover, this research identifies key research directions and offers important recommendations to promote the responsible deployment of T2V models.

Robot grasping and manipulation represent pivotal aspects of robotics research with profound implications for the future of autonomous systems. This report delves into the intricacies of designing robotic hands, the hurdles in creating robust manipulation actions, and the advancements in the field that poised to catalyze a new era of autonomy. Drawing inspiration from science fiction's portrayal of robotics, we bridge the conceptual gap between fiction and ongoing real-world technical research, aiming to provide a comprehensive overview for students interested in robotics.

This paper proposes a comprehensive 3GPPcnaompatible channel model with statistical enhancement tailored for indoor short-range device-to-device (D2D) communications operating in the frequency range (FR) of 52.6-71.0 GHz termed FR2-2. Regardless of the existence of various channel models at this band for indoor communications, there will be a need for developing a channel model compatible with and understandable from the current 3GPP stochastic channel model (SCM) to facilitate the discussion in the 3GPP for developing such FR2-2 short-range D2D communication framework based on the fifthgeneration (5G) new radio (NR). Indeed, such a futuristic vision can be foreseen from the fact that the 3GPP is discussing the evolution of sidelink, referred to as a D2D communication framework; however, there are no 3GPP SCM-compatible channel models applicable to FR2-2 short-range D2D communications. To fill this void, we propose the channel model coined 3GPPCompFR2-InS that allows us to generate channel impulse responses (CIR) for computer simulations, which is suitable for various indoor short-range D2D communication scenarios while retailing the similarity in terms of the implementation policy of the 3GPP SCM. 3GPPCompFR2-InS is verified based on the real-world measurements at the 60 GHz band from the viewpoint of both the validity of the channel model parameters and that of the statistical behavior of the generated CIRs.This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

Stress-induced progressive deformations in fractured rocks with increasing effective pressure generally undergo nonlinear elastic (due to the closure of compliant pores), hyperelastic (due to residual stress), and inelastic (due to fracture growth) deformations prior to mechanical failure. Wave propagation in such rocks involves the complex interaction of fracture-and stress-induced changes in both velocity and anisotropy. With attention to nonlinear elastic and hyperelastic deformations, we incorporate acoustoelasticity into the traditional Hudson/Cheng models to describe the coupling of fracture-induced and stress-induced anisotropies. The resulting acoustoelastic Hudson model (AHM) is valid for the crack density smaller than 0.1 whereas the Padé AHM could handle higher crack densities. We extend the Padé AHM to consider the stress-induced crack closure with nonlinear elastic deformations by incorporating the dual-porosity model. These models approach the coupled anisotropies with different accuracies and computational complexities. The plane-wave analyses and effective-moduli calculations of stressed fractured rocks with varying crack densities determine the accuracy of these models under the isotropic (confining) and anisotropic (uniaxial and pure shear) prestress conditions. The relevant Thomsen parameters are applied to experimental data to validate the applicability. Finite-difference simulations are implemented to identify the contribution of different anisotropies through the variety of wavefronts, depending on fracture orientation, crack density, prestress mode and magnitude, and loading direction. Particular attention is paid to the anisotropic prestress, where the coupled anisotropies are constructive or destructive interference, strongly related to the relativity between fracture strike and loading direction. The stress-induced crack closure will reduce the fracture anisotropy so that the stress-induced background anisotropy dominates the shape of wavefronts with increasing prestress.

In this paper, we analyze the secrecy performance of a two-hop cooperative network consisting solely of energy-harvesting self-sustaining nodes drawing energy from a multi-antenna power beacon (PB). Performance of such networks is quite different from that with powered nodes. We consider optimal combining of the direct and relayed signals at the multi-antenna destination as well as the multi-antenna eavesdropper. Since availability of channel state information at the source is impractical in such networks, we assume fixed-rate signaling. To implement incremental signaling, we utilize feedback bits from the destination. Assuming practical nonlinear EH, exact and approximate expressions are derived for the secrecy outage probability of the selective decode-and-forward (SDF) and the incremental decode-and-forward (IDF) relaying schemes. It is demonstrated that IDF has much better secrecy performance than SDF just as with powered nodes. However, unlike with powered nodes, the secrecy performance is a convex function of the transmit power of PB. We propose a novel power back-off scheme to improve secrecy under different network operating conditions. The security-reliability trade-off (SRT) is analyzed to highlight the trade-off between outage and secrecy performance with the power back-off scheme. Simulation results validate the analytical expressions.

Dynamic Spectrum Access (DSA) is a key mechanism for meeting the ever-increasing demand for emerging wireless services. DSA involves managing and assigning available spectrum resources in a way that minimizes interference and allows RF coexistence between heterogeneous devices and systems. Spectrum Consumption Models (SCMs)-defined in the IEEE 1900.5.2 standard, offer a mechanism for RF devices to: (i) declare the characteristics of their intended spectrum use and their interference protection needs; and (ii) determine compatibility (non-interference) with existing devices. In this paper, we propose a novel SCM-based Spectrum Deconfliction (SD) algorithm that dynamically configures RF operational parameters (e.g., center frequency and transmission power) of a target transmitter-receiver pair aiming to minimize interference with existing devices/systems. We also propose sequential and distributed DSA methods that use the SD algorithm for assigning spectrum in large-scale networks. To evaluate the performance of our methods in terms of computation time, spectrum assignment efficiency, and overhead, we use two custom-made simulation platforms. Finally, to experimentally demonstrate the feasibility of our methods, we build a proof-of-concept implementation in the NSF PAWR COSMOS wireless testbed. The results reveal the advantages of using SCMs and their capabilities to conduct spectrum assignments in dynamic and congested communication environments.

With the advent of Healthcare 4.0, there is increased interest from researchers the world over in the application of modern, cutting-edge Artificial Intelligence (AI) and Quantum Artificial Intelligence (QAI) algorithms in solving healthcare challenges. The era of Quantum Computing (QC) promises to bring significant advancements in several areas of healthcare such that it may be sensible to give this hybrid Quantum/Classical paradigm its own name-Healthcare4Q. The potential of QC will extend the reach of Healthcare4Q with the help of diverse technologies such as quantum-enabled wearables, quantum-secure transfer and storage of data, and quantum computing at edge, fog, and cloud. All of these technologies promise to catapult Healthcare4Q to become the most capable healthcare framework in the advancement of medical innovations and improvement of patient care. An integral part of a person's health lies in cardiovascular health, and thus prioritizing and optimizing cardiovascular health remains vital to the broader goals of public health and healthcare sustainability. In this study, under the paradigm of Healthcare4Q, we propose a framework called the Quantum AIdriven Heart Health Framework (QAIHHF) that can provide advanced predictive intelligence to healthcare providers by utilizing historical and real-time data and processing capabilities proposed in Healthcare4Q. We show that when applied to various diagnostics and health indicators such as ECG data, the Quantum AI provides accuracy at a level equal to or higher as compared to the classical methods thus proving itself to be the critical component that will herald the era of Healthcare4Q.

Low-power non-volatile memories operating down to deep cryogenic temperatures are important for a large spectrum of applications from high-performance computing and electronics interfacing quantum computing hardware to space-based electronics. Despite the potential of Hf0.5Zr0.5O2 (HZO), thanks to its compatibility with complementary metal-oxide-semiconductor (CMOS) back-end-of line processing, only a few studies of HZO based memory devices down to cryogenic operation temperatures exist. Here, we report on analog ferroelectric memory stack fabrication with 10 nm HZO and its detailed characterization under a wide range of pulse amplitudes and frequencies from 300 K down to 4 K. When operated at temperatures below 100 K, HZO devices can support high amplitude voltage pulses, yielding record high remnant polarization Pr of up to 75µC cm-2 at ±7 Vp (14 Vpp) pulse amplitudes accompanied with frequency-dependent memory window between 6 and 8 V. Devices show excellent endurance exceeding 109 cycles of ±5 Vp (10 Vpp) and Pr of 30 µC cm-2 without significant degradation of coercive voltages or loss of polarization at cryogenic temperatures. At least 20 reproducible analog states for temperatures below 100 K with almost ideal linearity of intermediate polarization states in both pulse directions are observed, demonstrating the high potential of analog cryogenic ferroelectric memories, essential for online training in In-Memory-Computing architecture.

This paper addresses the optimization of container unloading and loading operations at ports, integrating quaycrane dual-cycling (QCDC) with dockyard rehandle minimization. We present a unified model encompassing both operations: ship container unloading and loading by quay crane, and the other is reducing dockyard rehandles while loading the ship. We recognize that optimizing one aspect in isolation can lead to suboptimal outcomes due to interdependencies. Specifically, optimizing unloading sequences for minimal operation time may inadvertently increase dockyard rehandles during loading and vice versa. To address this NP-hard problem, we propose a hybrid genetic algorithm (GA) QCDC-DR-GA comprising 1dimensional and 2-dimensional GA components. Our model, QCDC-DR-GA, consistently outperforms four state-of-the-art methods in maximizing dual cycles and minimizing dockyard rehandles. Compared to those methods, it reduced 15-20% of total operation time for large vessels. Results underscore the inefficiency of separately optimizing QCDC and dockyard rehandles. Fragmented approaches, such as QCDC Scheduling Optimized by bi-level GA and GA-ILSRS (Scenario 2), show limited improvement compared to QCDC-DR-GA. As in GA-ILSRS (Scenario 1), neglecting dual-cycle optimization leads to inferior performance than our proposed QCDC-DR-GA.

This letter proposes a passive matrix addressing method to dramatically reduce the number of controlling (biasing) lines for large-scale beam-steering arrays. Phase shifting of each radiation element is accomplished in two stages by the row and column biasing lines located at two stacked layers, which offer a pair of orthogonal wave vectors. For an antenna array with M rows and N columns, conventional direct addressing designs need M * N biasing lines, while the proposed method needs only M+N biasing lines to complete the full-space beam-steering. The proposed method simplifies the routing of control lines and the design of driving circuits. Both theoretical and numerical analyses are provided to validate this method.