Papers

Speaker embeddings are widely used in speaker verification systems and other applications where it is useful to characterise the voice of a speaker with a fixed-length vector. These embeddings tend to be treated as "black box" encodings, and how they relate to conventional acoustic and phonetic dimensions of voices has not been widely studied. In this paper we investigate how state-of-the-art speaker embedding systems represent the acoustic characteristics of speakers as described by conventional acoustic descriptors, age, and gender. Using a large corpus of 10,000 speakers and three embedding systems we show that a small set of 9 acoustic parameters chosen to be "interpretable" predict embeddings about the same as 7 principal components, corresponding to over 50% of variance in the data. We show that some principal dimensions operate differently for male and female speakers, suggesting there is implicit gender recognition within the embedding systems. However we show that speaker age is not well captured by embeddings, suggesting opportunities exist for improvements in their calculation.

We demonstrate digital backpropagation-based compensation of fibre nonlinearities in the near-zero dispersion regime of the O-band. Single-step DBP effectively mitigates self-phase modulation, achieving SNR gains of up to 1.6 dB for 50 Gbaud PDM-256QAM transmission over a 2-span 151 km SMF-28 ULL fibre link.

This paper presents a framework for stabilizing nonlinear systems in state-dependent form. We first develop a model-based formulation that constructs a polytope enclosing the state-dependent system matrices over a prescribed state region and then synthesizes the controller by solving linear matrix inequalities (LMIs) at the polytope vertices. To establish stability and region of attraction (ROA) guarantees from the vertex LMIs, we develop a model-based mechanism and then apply it to direct data-driven control. In the data-driven setting, we use a known library of basis functions to represent the dynamics with unknown coefficients. A polytope captures the basis-function variation over a state region, and a data-consistent set characterizes the noise-induced coefficient uncertainty. This data-consistent set is incorporated directly into the vertex LMIs via Petersen's lemma. Building on our model-based mechanism, the resulting vertex LMIs yield a controller that guarantees stability and provides a certified ROA. The data-driven formulation further establishes disturbance-robustness guarantees and ROA estimates under input saturation. Numerical simulations and physical experiments demonstrate the effectiveness of the proposed controller.

Who is being described, where are they, and what are they doing? Referring Atomic Video Action Recognition (RAVAR) answers these questions jointly by grounding a natural-language reference to a person and recognizing that person's fine-grained atomic actions in complex multi-person videos. Progress in RAVAR is constrained by limited benchmark scale and weak alignment between fine-grained linguistic and scene cues and temporally coherent visual cues. We address both challenges with a new dataset and model. We introduce RefAVA++, a large-scale dataset comprising 2,950,920$ frames, 75,111 annotated person instances, and 80 atomic action categories. Its references describe appearance and spatial attributes while deliberately omitting action labels, requiring models to infer actions directly from visual cues. We further propose RefAtomNet++, which models complementary semantics at the holistic-sentence, partial-keyword, and scene-attribute levels. Fine-grained semantic cues retrieve aligned visual tokens across time to construct trajectories, which are aggregated through Mamba-based state-space modeling and fused using multi-hierarchical semantic-aligned cross-attention. This design enables accurate joint person localization and multi-label atomic action recognition. Equipped with the InternVideo2.5 backbone, RefAtomNet++ achieves 50.39%/51.14% mIoU, 59.21%/59.55% mAP, and 73.89%/76.59% AUROC on RefAVA, and 49.49%/50.33% mIoU, 62.33%/61.90% mAP, and 76.06%/76.10% AUROC on RefAVA++ validation/test sets, respectively. The dataset and code are available at https://github.com/KPeng9510/refAVA2.

Audio-visual speech enhancement (AVSE) has been found to be particularly useful at low signal-to-noise (SNR) ratios due to the immunity of the visual features to acoustic noise. However, a significant gap exists in AVSE methods tailored to enhance spatial audio under low-SNR conditions. The latter is of growing interest with augmented reality applications. To address this gap, we present a multi-channel AVSE framework based on VisualVoice that leverages spatial cues from microphone arrays and visual information for enhancing the target speaker in noisy environments. We also introduce MAVe, a novel database containing multi-channel audio-visual signals in controlled, reproducible room conditions across a wide range of SNR levels. Experiments demonstrate that the proposed method consistently achieves significant gains in SI-SDR, STOI, and PESQ, particularly in low SNRs. Binaural signal analysis further confirms the preservation of spatial cues and intelligibility.

Effective image deblurring typically relies on large and fully paired datasets of blurred and corresponding sharp images. However, obtaining such accurately aligned data in the real world poses a number of difficulties, limiting the effectiveness and generalizability of existing deblurring methods. To address this scarcity of data dependency, we present a novel dictionary learning based deblurring approach for jointly estimating a structured blur matrix and a high resolution image dictionary. This framework enables robust image deblurring across different degrees of data supervision. Our method is thoroughly evaluated across three distinct experimental settings: (i) full supervision involving paired data with explicit correspondence, (ii) partial supervision employing unpaired data with implicit relationships, and (iii) unsupervised learning using non-correspondence data where direct pairings are absent. Extensive experimental validation, performed on synthetically blurred subsets of the CMU-Cornell iCoseg dataset and the real-world FocusPath dataset, consistently shows that the proposed framework has superior performance compared to conventional coupled dictionary learning approaches. The results validate that our approach provides an efficient and robust solution for image deblurring in data-constrained scenarios by enabling accurate blur modeling and adaptive dictionary representation with a notably smaller number of training samples.

In the Industrial Internet of Things (IIoT), the frequent transmission of large amounts of data over wireless networks should meet the stringent timeliness requirements. Particularly, the freshness of packet status updates has a significant impact on the system performance. In this paper, we propose an age-of-information (AoI)-aware multi-base station (BS) real-time monitoring framework to support extensive IIoT deployments. To meet the freshness requirements of IIoT, we formulate a joint task offloading and resource allocation optimization problem with the goal of minimizing long-term average AoI. Tackling the core challenges of combinatorial explosion in multi-BS decision spaces and the stochastic dynamics of IIoT systems is crucial, as these factors render traditional optimization methods intractable. Firstly, an innovative branching-based Dueling Double Deep Q-Network (Branching-D3QN) algorithm is proposed to effectively implement task offloading, which optimizes the convergence performance by reducing the action space complexity from exponential to linear levels. Then, an efficient optimization solution to resource allocation is proposed by proving the semi-definite property of the Hessian matrix of bandwidth and computation resources. Finally, we propose an iterative optimization algorithm for efficient joint task offloading and resource allocation to achieve optimal average AoI performance. Extensive simulations demonstrate that our proposed Branching-D3QN algorithm outperforms both state-of-the-art DRL methods and classical heuristics, achieving up to a 75% enhanced convergence speed and at least a 22% reduction in the long-term average AoI.

The secure operation of power distribution systems is challenged by the growing integration of distributed energy resources. Leveraging the flexibility of battery storage offers a cost-effective alternative to measures like generation curtailment, which results in energy losses. However, developing an effective operational model for battery storage is hindered by inaccurate grid models, unavailability of load data, nonlinear relationship between power injections and network states, intertemporal constraints, and complex electrochemical and thermal dynamics. To address these challenges, this paper proposes a data-driven operational control scheme for battery storage in distribution systems. Linear and convex quadratic operational constraints are constructed based on real-time distribution system and battery storage measurements. Lyapunov optimization decouples multi-period battery operation, enabling a real-time, forecast-free control strategy with low computational complexity. Numerical studies using nonlinear distribution system and battery storage simulators validate the effectiveness of the approach in ensuring secure distribution system operation and satisfaction of voltage and thermal constraints of battery storage.

This work considers the iterative solution of large-scale problems subject to non-symmetric matrices or operators arising in discretizations of (port-)Hamiltonian partial differential equations. We consider problems governed by an operator $\mathcal{A}=\mathcal{H}+\mathcal{S}$ with symmetric part $\mathcal{H}$ that is positive (semi-)definite and skew-symmetric part $\mathcal{S}$. Prior work has shown that the structure and sparsity of the associated linear system enables Krylov subspace solvers such as the generalized minimal residual method (GMRES) or short recurrence variants such as Widlund's or Rapoport's method using the symmetric part $\mathcal{H}$, or an approximation of it, as preconditioner. In this work, we analyze the resulting condition numbers, which are crucial for fast convergence of these methods, for various partial differential equations (PDEs) arising in diffusion phenomena, fluid dynamics, and elasticity. We show that preconditioning with the symmetric part leads to a condition number uniform in the mesh size in case of elliptic and parabolic PDEs where $\mathcal{H}^{-1}\mathcal{S}$ is a bounded operator. Further, we employ the tailored Krylov subspace methods in optimal control by means of a condensing approach and a constraint preconditioner for the optimality system. We illustrate the results by various large-scale numerical examples and discuss efficient evaluations of the preconditioner, such as incomplete Cholesky factorization or the algebraic multigrid method.

The channel state information (CSI) of an eavesdropper is crucial for physical layer security (PLS) design, but it is difficult to obtain due to the passive and non-cooperative nature of the eavesdropper. To this end, integrated sensing and communication (ISAC) offers a novel solution by estimating the CSI of the eavesdropper based on sensing information. However, existing studies normally impose explicit and fixed sensing performance requirement without considering the varying communication conditions, which hinders the system from fully exploiting the synergy between sensing and communication. To address this issue, this paper proposes sensing-enhanced secure communication with adaptive sensing performance. Specifically, we formulate the sensing performance implicitly in the information leakage rate and adaptively optimize it for the minimization of the power consumption, offering enhanced flexibility and adaptability in sensing performance. We consider both centralized and decentralized designs to thoroughly investigate the impact of network structure on system performance and complexity. Specifically, we devise a block coordinate descent (BCD)-based method for centralized design. For decentralized design, we develop an optimization framework based on consensus alternating direction method of multipliers (ADMM) to reduce complexity and information exchange overhead. Experimental results demonstrate the advantage of the proposed implicit sensing performance requirement design due to its capability to adaptively adjust the sensing performance to enhance the system performance for varying system configurations.

Synthetic Aperture Radar (SAR) image captioning enables scene-level semantic understanding and plays a crucial role in applications such as military intelligence and urban planning, but its development is limited by the scarcity of high-quality datasets. To address this, we present FSAR-Cap, a large-scale SAR captioning dataset with 14,480 images and 72,400 image-text pairs. FSAR-Cap is built on the FAIR-CSAR detection dataset and constructed through a two-stage annotation strategy that combines hierarchical template-based representation, manual verification and supplementation, prompt standardization. Compared with existing resources, FSAR-Cap provides richer fine-grained annotations, broader category coverage, and higher annotation quality. Benchmarking with multiple encoder-decoder architectures verifies its effectiveness, establishing a foundation for future research in SAR captioning and intelligent image interpretation.

Digital twin (DT) is a core enabler of sixth generation (6G) mobile systems. As a prerequisite for DT, scatterer geometric reconstruction (SGR) in propagation environments is essential but typically requires extra sensors such as cameras and LiDAR. With integrated sensing and communication (ISAC) in 6G, we reinterpret the linear sampling method (LSM) from a wireless channel viewpoint and propose a CSI based variant for sensor free SGR: by exploiting the shared channel characteristics of multipath and scattering, in band CSI replaces the scattered field measurements usually required by LSM. However, aperture limited arrays reduce LSM robustness. To address this, we propose matched filtering enhanced multi frequency LSM (MF MLSM). Multi frequency data increases frequency diversity, and matched filtering coherently aligns inter frequency phases to avoid artifacts, both of which improve robustness. Experiments with apertures of 93.6 deg, 144 deg, and 180 deg and SNRs of 27 dB and 12 dB demonstrate robust SGR with this approach.

In this paper, we explore the untapped potential of Whisper, a well-established automatic speech recognition (ASR) foundation model, in the context of L2 spoken language assessment (SLA). Unlike prior studies that extrinsically analyze transcriptions produced by Whisper, our approach goes a step further to probe its latent capabilities by extracting acoustic and linguistic features from hidden representations. With only a lightweight classifier being trained on top of Whisper's intermediate and final outputs, our method achieves strong performance on the GEPT picture-description dataset, outperforming existing cutting-edge baselines, including a multimodal approach. Furthermore, by incorporating image and text-prompt information as auxiliary relevance cues, we demonstrate additional performance gains. Finally, we conduct an in-depth analysis of Whisper's embeddings, which reveals that, even without task-specific fine-tuning, the model intrinsically encodes both ordinal proficiency patterns and semantic aspects of speech, highlighting its potential as a powerful foundation for SLA and other spoken language understanding tasks.

Conformal Prediction (CP) is a powerful statistical machine learning tool to construct uncertainty sets with coverage guarantees, which has fueled its extensive adoption in generating prediction regions for decision-making tasks, e.g., Trajectory Optimization (TO) in uncertain environments. However, existing methods predominantly employ a sequential scheme, where decisions rely unidirectionally on the prediction regions, and consequently the information from decision-making fails to be fed back to instruct CP. In this paper, we propose a novel Feedback-Based CP (Fb-CP) framework for shrinking-horizon TO with a joint risk constraint over the entire mission time. Specifically, a CP-based posterior risk calculation method is developed by fully leveraging the realized trajectories to adjust the posterior allowable risk, which is then allocated to future times to update prediction regions. In this way, the information in the realized trajectories is continuously fed back to the CP, enabling attractive feedback-based adjustments of the prediction regions and a provable online improvement in trajectory performance. Furthermore, we theoretically prove that such adjustments consistently maintain the coverage guarantees of the prediction regions, thereby ensuring provable safety. Additionally, we develop a decision-focused iterative risk allocation algorithm with theoretical convergence analysis for allocating the posterior allowable risk which closely aligns with Fb-CP. Furthermore, we extend the proposed method to handle distribution shift. The effectiveness and superiority of the proposed method are demonstrated through benchmark experiments.

High sound pressure levels (SPL) pose notable risks in loud environments, particularly due to noise-induced hearing loss. Ill-fitting earplugs often lead to sound leakage, a phenomenon this study seeks to investigate. To validate our methodology, we first obtained computational and experimental acoustic transmission data for stand-alone slit resonators and orifices, for which extensive published data are readily available for comparison. We then examined the frequency-dependent acoustic power absorption coefficient and transmission loss (TL) across various leakage geometries, modeled using different orifice diameters. Experimental approaches spanned a frequency range of 1–5 kHz under SPL conditions of 120–150 dB. Key findings reveal that unsealed silicone rubber earplugs demonstrate an average TL reduction of approximately 18 dB at an overall incident SPL (OISPL) of 120 dB. Direct numerical simulations further highlight SPL-dependent acoustic dissipation mechanisms, showing the conversion of acoustic energy into vorticity in ill-fitting earplug models at an OISPL of 150 dB. These results highlight the role of earplug design for high-sound-pressure-level environments.

This research investigates designing a supervisory feedback controller for a hybrid power plant that coordinates the wind, solar, and battery energy storage plants to meet the desired power demands. We have explored an online feedback control design that does not require detailed knowledge about the models, known as feedback optimization. The control inputs are updated using the gradient information of the cost and the outputs with respect to the input control commands. This enables us to adjust the active power references of wind, solar, and storage plants to meet the power generation requirements set by grid operators. The methodology also ensures robust control performance in the presence of uncertainties in the weather. In this paper, we focus on describing the supervisory feedback optimization formulation and control-oriented modeling for individual renewable and storage components of the hybrid power plant. The proposed supervisory control has been integrated with the hybrid plant co-simulation engine, Hercules, demonstrating its effectiveness in more realistic simulation scenarios.

Current spinal pain management procedures, such as radiofrequency ablation (RFA) and epidural steroid injection (ESI), rely on fluoroscopy for needle placement which exposes patients and physicians to ionizing radiation. In this paper, we investigate a radiation-free surgical navigation system for spinal pain management procedures that combines magnetic resonance imaging (MRI) with fiducial ArUco marker-based augmented reality (AR). High-resolution MRI scans of a lumbar spinal phantom were obtained and assembled as a surface mesh. Laplacian smoothing algorithms were then applied to smoothen the surface and improve the model fidelity. A commercially available stereo camera (ZED2) was used to track single or dual fiducial ArUco markers on the patient to determine the patient's real-time pose. Custom AR software was applied to overlay the MRI image onto the patient, allowing the physician to see not only the outer surface of the patient but also the complete anatomy of the patient below the surface. Needle-insertion trials on a 3D-printed 3-vertebra phantom showed that dual-ArUco marker tracking increased the accuracy of needle insertions and reduced the average needle misplacement distance compared to single-ArUco marker procedures. The average needle misplacement is comparable to the average deviation of 2 mm for conventional epidural techniques using fluoroscopy. Our radiation-free system demonstrates promise to serve as an alternative to fluoroscopy by improving image-guided spinal navigation.

Algorithm unrolling methods have proven powerful for solving the regularized least squares problem in computational magnetic resonance imaging (MRI). These approaches unfold an iterative algorithm with a fixed number of iterations, typically alternating between a neural network-based proximal operator for regularization, a data fidelity operation and auxiliary updates with learnable parameters. While the connection to optimization methods dictate that the proximal operator network should be shared across unrolls, this can introduce artifacts or blurring. Heuristically, practitioners have shown that using distinct networks may be beneficial, but this significantly increases the number of learnable parameters, making it challenging to prevent overfitting. To address these shortcomings, by taking inspirations from proximal operators with varying thresholds in approximate message passing (AMP) and the success of time-embedding in diffusion models, we propose a time-embedded algorithm unrolling scheme for inverse problems. Specifically, we introduce a novel perspective on the iteration-dependent proximal operation in vector AMP (VAMP) and the subsequent Onsager correction in the context of algorithm unrolling, framing them as a time-embedded neural network. Similarly, the scalar weights in the data fidelity operation and its associated Onsager correction are cast as time-dependent learnable parameters. Our extensive experiments on the fastMRI dataset, spanning various acceleration rates and datasets, demonstrate that our method effectively reduces aliasing artifacts and mitigates noise amplification, achieving state-of-the-art performance. Furthermore, we show that our time-embedding strategy extends to existing algorithm unrolling approaches, enhancing reconstruction quality without increasing the computational complexity significantly.

The study of complex systems has attracted widespread attention from researchers in the fields of natural sciences, social sciences, and engineering. Prediction is one of the central issues in this field. Although most related studies have focused on prediction methods, research on the predictability of complex systems has received increasing attention across disciplines–aiming to provide theories and tools to address a key question: What are the limits of prediction accuracy? Predictability itself can serve as an important feature for characterizing complex systems, and accurate estimation of predictability can provide a benchmark for the study of prediction algorithms. This allows researchers to clearly identify the gap between current prediction accuracy and theoretical limits, thereby helping them determine whether there is still significant room to improve existing algorithms. More importantly, investigating predictability often requires the development of new theories and methods, which can further inspire the design of more effective algorithms. Over the past few decades, this field has undergone significant evolution. In particular, the rapid development of data science has introduced a wealth of data-driven approaches for understanding and quantifying predictability. This review summarizes representative achievements, integrating both data-driven and mechanistic perspectives. After a brief introduction to the significance of the topic in focus, we will explore three core aspects: the predictability of time series, the predictability of network structures, and the predictability of dynamical processes. Finally, we will provide extensive application examples across various fields and outline open challenges for future research.

Lung cancer, a malignancy originating in lung tissues, is commonly diagnosed and classified using medical imaging techniques, particularly computed tomography (CT). Despite the integration of machine learning and deep learning methods, the predictive efficacy of automated systems for lung cancer classification from CT images remains below the desired threshold for clinical adoption. Existing research predominantly focuses on binary classification, distinguishing between malignant and benign lung nodules. In this study, a novel deep learning-based approach is introduced, aimed at an improved multi-class classification, discerning various subtypes of lung cancer from CT images. Leveraging a pre-trained ResNet model, lung tissue images were classified into three distinct classes, two of which denote malignancy and one benign. Employing a dataset comprising 15,000 lung CT images sourced from the LC25000 histopathological images, the ResNet50 model was trained on 10,200 images, validated on 2,550 images, and tested on the remaining 2,250 images. Through the incorporation of custom layers atop the ResNet architecture and meticulous hyperparameter fine-tuning, a remarkable test accuracy of 98.8% was recorded. This represents a notable enhancement over the performance of prior models on the same dataset.

The high penetration of distributed energy resources, resulting in backfeed of power at the transmission and distribution interface, is causing conventional underfrequency load shedding (UFLS) schemes to become nonconforming. Adaptive schemes that update UFLS relay settings recursively in time offer a solution, but existing adaptive techniques that obtain UFLS relay settings with linearized or reduced-order model formulations fail to capture AC nonlinear network behavior. In practice, this will result in relays unable to restore system frequency during adverse disturbances. We formulate an adaptive UFLS problem as a trajectory optimization and include the full AC nonlinear network dynamics to ensure AC feasibility and time-coordinated control actions. We include binary decisions to model relay switching action and time-delayed multi-stage load-shedding. However, this formulation results in an intractable MINLP problem. To enforce model tractability, we relax these binary variables into continuous surrogates and reformulate the MINLP as a sequence of NLPs. We solve the NLPs with a homotopy-driven method that enforces near-integer-feasible solutions. We evaluate the framework on multiple synthetic transmission systems and demonstrate that it scales efficiently to networks exceeding 1500+ nodes with over 170k+ continuous and 73k+ binary decision variables, while successfully recovering binary-feasible solutions that arrest the frequency decline during worst-case disturbance.

This letter proposes a novel pinching antenna systems (PASS) enabled non-orthogonal multiple access (NOMA) multi-access edge computing (MEC) framework. An optimization problem is formulated to minimize the maximum task delay by optimizing offloading ratios, transmit powers, and pinching antenna (PA) positions, subject to constraints on maximum transmit power, user energy budgets, and minimum PA separation to mitigate coupling effects. To address the non-convex problem, a bisection search-based alternating optimization (AO) algorithm is developed, where each subproblem is iteratively solved for a given task delay. Numerical simulations demonstrate that the proposed framework significantly reduces the task delay compared to benchmark schemes.

The rapid evolution of modern manufacturing systems is driven by the integration of emerging metaverse technologies such as artificial intelligence (AI), digital twin (DT) with different forms of extended reality (XR) like virtual reality (VR), augmented reality (AR), and mixed reality (MR). These advances confront manufacturing workers with complex and evolving environments that demand digital literacy for problem solving in the future workplace. However, manufacturing industry faces a critical shortage of skilled workforce with digital literacy in the world. Further, global pandemic has significantly changed how people work and collaborate digitally and remotely. There is an urgent need to rethink digital platformization and leverage emerging technologies to propel industrial evolution toward human-centered manufacturing metaverse (MfgVerse). This paper presents a forward-looking perspective on the development of smart MfgVerse, highlighting current efforts in learning factory, cognitive digital twinning, and the new sharing economy of manufacturing-as-a-service (MaaS). MfgVerse is converging into multiplex networks, including a social network of human stakeholders, an interconnected network of manufacturing things or agents (e.g., machines, robots, facilities, material handling systems), a network of digital twins of physical things, as well as auxiliary networks of sales, supply chain, logistics, and remanufacturing systems. We also showcase the design and development of a learning factory for workforce training in extended reality. Finally, future directions, challenges, and opportunities are discussed for human-centered manufacturing metaverse. We hope this work helps stimulate more comprehensive studies and in-depth research efforts to advance MfgVerse technologies.

Discrete representation learning has shown promising results across various domains, including generation and understanding in image, speech and language. Inspired by these advances, we propose MuseTok, a tokenization method for symbolic music, and investigate its effectiveness in both music generation and understanding tasks. MuseTok employs the residual vector quantized-variational autoencoder (RQ-VAE) on bar-wise music segments within a Transformer-based encoder-decoder framework, producing music codes that achieve high-fidelity music reconstruction and accurate understanding of music theory. For comprehensive evaluation, we apply MuseTok to music generation and semantic understanding tasks, including melody extraction, chord recognition, and emotion recognition. Models incorporating MuseTok outperform previous representation learning baselines in semantic understanding while maintaining comparable performance in content generation. Furthermore, qualitative analyses on MuseTok codes, using ground-truth categories and synthetic datasets, reveal that MuseTok effectively captures underlying musical concepts from large music collections.

Artificial transmission lines built with lumped-element inductors and capacitors form the backbone of broadband, nearly quantum-limited traveling-wave parametric amplifiers (TWPAs). When tailoring these transmission lines for parametric processes, nonlinear elements are added, typically nonlinear inductances in superconducting circuits, and energy and momentum conservation between interacting tones must be enforced through careful design of the ATL dispersion relation. However, a unified theoretical framework describing achievable dispersion relations is lacking. Here, I develop such a framework, borrowing from periodic structure theory and passive network synthesis. These complementary approaches divide the design space: periodic loading synthesis employs spatial modulation of frequency-independent components, while filter synthesis employs frequency-dependent responses in spatially-uniform components. The framework reveals fundamental constraints and enables the discovery of novel TWPA architectures. In particular, I design a kinetic inductance TWPA with a novel phase-matching architecture, and a backward-pumped Josephson TWPA exploiting an ambidextrous i.e., right-left-handed transmission line.

Wolff-Parkinson-White (WPW) syndrome, a congenital cardiac conduction abnormality with low prevalence, carries a significant risk of sudden cardiac death. Early identification remains challenging due to screening costs and professional resource scarcity. This retrospective real-world study systematically evaluates an integrated Artificial Intelligence-enabled mobile screening system comprising portable single-lead devices, AI primary screening, and cardiologist review. Analyzing 3,566,626 ECG records from 87,836 individuals between 2019 and 2025, the AI model achieved an AUC of 0.6676 and a specificity of 95.92% in complex real-world signal environments. Despite predictive probability bias inherent in ultra-low prevalence contexts, the model demonstrated stable risk stratification, with high-confidence scores concentrated among true positive individuals. The risk of detecting WPW in AI-positive records was 86.2-fold higher than in AI-negative records. By implementing a human-AI collaborative workflow, the volume of ECGs requiring manual review was reduced by approximately 99.5% compared to universal screening. In an ideal collaborative scenario, an average of only 18 ECGs required review to confirm one WPW case, representing a more than 60-fold increase in screening efficiency. Compared to traditional 12-lead ECGs and electrophysiological studies, this system significantly reduced time and medical costs. Our findings suggest that a risk-stratification-based human-AI collaborative system provides a promising paradigm for the early public health detection of low-prevalence, high-risk arrhythmias.

In facial image generation, current text-to-image models often suffer from facial attribute leakage and insufficient physical consistency when responding to local semantic instructions. In this study, we propose Face-MakeUpV2, a facial image generation model that aims to maintain the consistency of face ID and physical characteristics with the reference image. First, we constructed a large-scale dataset FaceCaptionMask-1M comprising approximately one million image-text-masks pairs that provide precise spatial supervision for the local semantic instructions. Second, we employed a general text-to-image pretrained model as the backbone and introduced two complementary facial information injection channels: a 3D facial rendering channel to incorporate the physical characteristics of the image and a global facial feature channel. Third, we formulated two optimization objectives for the supervised learning of our model: semantic alignment in the model's embedding space to mitigate the attribute leakage problem and perceptual loss on facial images to preserve ID consistency. Extensive experiments demonstrated that our Face-MakeUpV2 achieves best overall performance in terms of preserving face ID and maintaining physical consistency of the reference images. These results highlight the practical potential of Face-MakeUpV2 for reliable and controllable facial editing in diverse applications.

Intelligent reflecting surface (IRS) assisted unmanned aerial vehicle (UAV) systems provide a new paradigm for reconfigurable and flexible wireless communications. To enable more energy efficient and spectrum efficient IRS assisted UAV wireless communications, this paper introduces a novel IRS-assisted UAV enabled spectrum sharing system with orthogonal frequency division multiplexing (OFDM). The goal is to maximize the energy efficiency (EE) of the secondary network by jointly optimizing the beamforming, subcarrier allocation, IRS phase shifts, and the UAV trajectory subject to practical transmit power and passive reflection constraints as well as UAV physical limitations. A physically grounded propulsion-energy model is adopted, with its tight upper bound used to form a tractable EE lower bound for the spectrum sharing system. To handle highly non convex, time coupled optimization problems with a mixed continuous and discrete policy space, we develop a deep reinforcement learning (DRL) approach based on the actor critic framework. Extended experiments show the significant EE improvement of the proposed DRL-based approach compared to several benchmark schemes, thus demonstrating the effectiveness and robustness of the proposed approach with mobility.

This article presents an overview and analysis of spatial-to-spectral harmonic-modulated arrays (SHAs). Compared to traditional analog or digital beamforming arrays, SHAs enable concurrent multi-beamforming without requiring substantial hardware replication. SHAs replace the need for hardware replication with frequency-domain multiplexing. Furthermore, SHAs have the potential to become key contributors to future 6G networks by enabling scalable multi-user communications, joint communication and sensing, and spatial interference mitigation. In addition, an analysis of the SHA's harmonic-modulation waveform and its effects on gain, noise and bandwidth is presented. A comb-like modulation waveform for SHAs that minimizes spectral inefficiency is proposed. Further, an analysis of the SHA's capability to independently steer multiple beams is presented. This capability is quantified in terms of the SHA's spatial-to-spectral degrees of freedom. Lastly, this work introduces a novel SHA architecture that provides three spatial-to-spectral degrees of freedom with minimal hardware replication.

A general and psychologically plausible collision avoidance driver model can improve transportation safety significantly. Most computational driver models found in the literature have used control theory methods only, and they are not established based on psychological theories. In this paper, a unified approach is presented based on concepts taken from psychology and control theory. The "task difficulty homeostasis theory", a prominent motivational theory, is combined with the "Lyapunov stability method" in control theory to present a general and psychologically plausible model. This approach is used to model driver steering behavior for collision avoidance. The performance of this model is measured by simulation of two collision avoidance scenarios at a wide range of speeds from 20 km/h to 170 km/h. The model is validated by experiments on a driving simulator. The results demonstrate that the model follows human behavior accurately with a mean error of 7 percent.