Papers

Brain lesion segmentation from multi-modal MRI often assumes fixed modality sets or predefined pathologies, making existing models difficult to adapt across cohorts and imaging protocols. Continual learning (CL) offers a natural solution but current approaches either impose a maximum modality configuration or suffer from severe forgetting in buffer-free settings. We introduce CLMU-Net, a replay-based CL framework for 3D brain lesion segmentation that supports arbitrary and variable modality combinations without requiring prior knowledge of the maximum set. A conceptually simple yet effective channel-inflation strategy maps any modality subset into a unified multi-channel representation, enabling a single model to operate across diverse datasets. To enrich inherently local 3D patch features, we incorporate lightweight domain-conditioned textual embeddings that provide global modality-disease context for each training case. Forgetting is further reduced through principled replay using a compact buffer composed of both prototypical and challenging samples. Experiments on five heterogeneous MRI brain datasets demonstrate that CLMU-Net consistently outperforms popular CL baselines. Notably, our method yields an average Dice score improvement of $\geq$ 18% while remaining robust under heterogeneous-modality conditions. These findings underscore the value of flexible modality handling, targeted replay, and global contextual cues for continual medical image segmentation. Our implementation is available at https://github.com/xmindflow/CLMU-Net.

Aortic valve opening (AO) events are crucial for detecting frequency and rhythm disorders, especially in real-world settings where seismocardiography (SCG) signals collected via consumer smartphones are subject to noise, motion artifacts, and variability caused by device heterogeneity. In this work, we present a robust deep-learning framework for SCG segmentation and rhythm analysis using accelerometer recordings obtained with consumer smartphones. We develop an enhanced U-Net v3 architecture that integrates multi-scale convolutions, residual connections, and attention gates, enabling reliable segmentation of noisy SCG signals. A dedicated post-processing pipeline converts probability masks into precise AO timestamps, whereas a novel adaptive 3D-to-1D projection method ensures robustness to arbitrary smartphone orientation. Experimental results demonstrate that the proposed method achieves consistently high accuracy and robustness across various device types and unsupervised data-collection conditions. Our approach enables practical, low-cost, and automated cardiac-rhythm monitoring using everyday mobile devices, paving the way for scalable, field-deployable cardiovascular assessment and future multimodal diagnostic systems.

Recent advances in singing voice synthesis (SVS) have attracted substantial attention from both academia and industry. With the advent of large language models and novel generative paradigms, producing controllable, high-fidelity singing voices has become an attainable goal. Yet the field still lacks a comprehensive survey that systematically analyzes deep-learning-based singing voice synthesis systems and their enabling technologies. To address the aforementioned issue, this survey first categorizes existing systems by task type and then organizes current architectures into two major paradigms: cascaded and end-to-end approaches. Moreover, we provide an in-depth analysis of core technologies, covering singing modeling and control techniques. Finally, we review relevant datasets, annotation tools, and evaluation benchmarks that support training and assessment. In appendix, we introduce training strategies and further discussion of SVS. This survey provides an up-to-date review of the literature on SVS models, which would be a useful reference for both researchers and engineers. Related materials are available at https://github.com/David-Pigeon/SyntheticSingers.

In this paper, we rigorously characterize for the first time the manifold of unitary and symmetric matrices, deriving its tangent space and its geodesics. The resulting parameterization of the geodesics (through a real and symmetric matrix) allows us to derive a new Riemannian manifold optimization (MO) algorithm whose most remarkable feature is that it does not need to set any adaptation parameter. We apply the proposed MO algorithm to maximize the achievable rate in a multiple-input multiple-output (MIMO) system assisted by a beyond-diagonal reconfigurable intelligent surface (BD-RIS), illustrating the method's performance through simulations. The MO algorithm achieves a significant reduction in computational cost compared to previous alternatives based on Takagi decomposition, while retaining global convergence to a stationary point of the cost function.

Active noise control (ANC) must adapt quickly when the acoustic environment changes, yet early performance is largely dictated by initialization. We address this with a Model-Agnostic Meta-Learning (MAML) co-initialization that jointly sets the control filter and the secondary-path model for FxLMS-based ANC while keeping the runtime algorithm unchanged. The initializer is pre-trained on a small set of measured paths using short two-phase inner loops that mimic identification followed by residual-noise reduction, and is applied by simply setting the learned initial coefficients. In an online secondary path modeling FxLMS testbed, it yields lower early-stage error, shorter time-to-target, reduced auxiliary-noise energy, and faster recovery after path changes than a baseline without re-initialization. The method provides a simple fast start for feedforward ANC under environment changes, requiring a small set of paths to pre-train.

Large Language Models (LLMs) such as ChatGPT promise revolutionary capabilities for Sixth-Generation (6G) wireless networks but their massive computational requirements and tendency to generate technically incorrect information create deployment barriers. In this work, we introduce MAINTAINED: autonomous artificial intelligence agent for wireless network deployment. Instead of encoding domain knowledge within model parameters, our approach orchestrates specialized computational tools for geographic analysis, signal propagation modeling, and network optimization. In a real-world case study, MAINTAINED outperforms state-of-the-art LLMs including ChatGPT-4o, Claude Sonnet 4, and DeepSeek-R1 by up to 100-fold in verified performance metrics while requiring less computational resources. This paradigm shift, moving from relying on parametric knowledge towards externalizing domain knowledge into verifiable computational tools, eliminates hallucination in technical specifications and enables edge-deployable Artificial Intelligence (AI) for wireless communications.

Wi-Fi networks increasingly suffer from performance degradation caused by contention-based channel access, dense deployments, and largely self-managed operation among mutually interfering access points (APs). In this paper, we propose a Digital Twin (DT) framework that captures the essential spatial and temporal characteristics of wireless channels and traffic patterns, enabling the prediction of likely future network scenarios while respecting physical constraints. Leveraging this predictive capability, we introduce two analytically derived performance upper bounds-one based on Shannon capacity and the other on latency behavior under CSMA-CA (Carrier Sense Multiple Access with Collision Avoidance)-that can be evaluated efficiently without time-consuming network simulations. By applying importance sampling to DT-generated scenarios, potentially risky network conditions can be identified within large stochastic scenario spaces. These same performance bounds are then used to proactively guide a gradient-based search for improved network configurations, with the objective of avoiding imminent performance degradation rather than pursuing globally optimal but fragile solutions. Simulation results demonstrate that the proposed approach can successfully predict time-dependent network congestion and mitigate it in advance, highlighting its potential for predictive and preventive Wi-Fi network management.

To address the issues of high operational costs and low energy efficiency (EE) caused by the dense deployment of small base stations (s-BSs) in 5G ultra-dense networks (UDNs), this paper first constructs a multi-objective mathematical optimization model targeting maximizing EE and minimizing the number of active BSs. The model incorporates key constraints including BS operational state, user equipment (UE)-BS connection relationship, and load threshold, laying a theoretical foundation for the coordinated optimization of energy conservation and quality of service. Based on this model, an integrated solution combining UE-BS initial connection optimization and load-sharing based BS sleeping is proposed. In the initial connection phase, with communication quality and BS load as dual constraints, efficient matching between UEs and optimal BSs is achieved through three sequential steps: communication feasibility screening, redundant connection removal, and overload load redistribution. This resolves the problems of load imbalance and difficult identification of redundant BSs in UDNs arising from unordered initial connections. In the BS sleeping phase, a BS sleeping index, comprehensively considering UE transferability and backup BS resources, is innovatively introduced to quantify BS dormancy priority. Through a closed-loop process involving low-load BS screening, adjacent BS load evaluation, and load sharing by two takeover BSs based on their capacity, accurate dormancy of redundant BSs and collaborative load migration are realized. Simulation results in a typical UDNs scenario demonstrate that, compared with the traditional baseline scheme, the proposed solution exhibits significant advantages in convergence speed, optimization of the number of active BSs, and EE improvement.

Multiple-input multiple-output (MIMO) systems require efficient and accurate channel estimation with low pilot overhead to unlock their full potential for high spectral and energy efficiency. While deep generative models have emerged as a powerful foundation for the channel estimation task, the existing approaches using diffusion-based and score-based models suffer from high computational runtime due to their stochastic and many-step iterative sampling. In this paper, we introduce a flow matching-based channel estimator to overcome this limitation. The proposed channel estimator is based on a deep neural network trained to learn the velocity field of wireless channels, which we then integrate into a plug-and-play proximal gradient descent (PnP-PGD) framework. Simulation results reveal that our formulated approach consistently outperforms existing state-of-the-art (SOTA) generative model-based estimators, achieves up to 49 times faster inference at test time, and reduces up to 20 times peak graphics processing unit (GPU) memory usage. Our code and models are publicly available to support reproducible research.

The dense concentration of low-altitude, slow-speed, and small-size targets in the complex low-altitude environment poses significant security challenges, including failures in continuous wide-area sensing and ambiguous target intent, which existing regulatory frameworks struggle to address. Integrated sensing and communication (ISAC), a hallmark of next-generation mobile communication, offers a transformative approach to low-altitude security governance. By leveraging existing cellular infrastructure and spectrum resources, ISAC enables the construction of a seamless wide-area sensing network, supports intelligent feature extraction and intent inference, facilitates real-time collaborative decision-making, and establishes a dynamic trust authentication framework. This article systematically reviews the technical system, analyzes the security challenges, forecasts the enabling value of ISAC, and discusses the resulting open problems and challenges, thereby laying a foundation for future research and industrial implementation.

Arabic spans over 30 spoken varieties, yet no open-source text-to-speech system unifies them. Key barriers include substantial cross-dialect lexical and phonological divergence, scarce synthesis-grade data, and the absence of a standardized multi-dialect evaluation benchmark. We present Habibi, a unified-dialectal Arabic TTS framework that addresses all three. Through a multi-step curation pipeline, we repurpose open-source ASR corpora into TTS training data covering 12+ regional dialects. A linguistically-informed curriculum learning strategy - progressing from Modern Standard Arabic to dialectal data - enables robust zero-shot synthesis without text diacritization. We further release the first standardized multi-dialect Arabic TTS benchmark, comprising over 11,000 utterances across 7 dialect subsets with manually verified transcripts. On this benchmark, our unified model matches or surpasses per-dialect specialized models. Both automatic metrics and human evaluations confirm that Habibi is highly competitive with ElevenLabs' Eleven v3 (alpha) in intelligibility, speaker similarity, and naturalness. Extensive ablations (~8,000 H100 GPU hours, 30+ configurations) validate each design choice. We open-source all checkpoints, training and inference code, and benchmark data - the first such release for multi-dialect Arabic TTS - at https://SWivid.github.io/Habibi/ .

In [1], a new modeling paradigm for developing rate-and-state-dependent, control-oriented friction models was introduced. The framework, termed Friction with Bristle Dynamics (FrBD), combines nonlinear analytical expressions for the friction coefficient with constitutive equations for bristle-like elements. Within the FrBD framework, this letter introduces two novel formulations based on the two most general linear viscoelastic models for solids: the Generalized Maxwell (GM) and Generalized Kelvin-Voigt (GKV) elements. Both are analyzed in terms of boundedness and passivity, revealing that these properties are satisfied for any physically meaningful parametrization. An application of passivity for control design is also illustrated, considering an example from robotics. The findings of this letter systematically integrate rate-and-state dynamic friction models with linear viscoelasticity.

Carbon credit systems have emerged as a policy tool to incentivize emission reductions and support the transition to clean energy. Reliable carbon-credit certification depends on mechanisms that connect actual, measured renewable-energy production to verifiable emission-reduction records. Although blockchain and IoT technologies have been applied to emission monitoring and trading, existing work offers limited support for certification processes, particularly for small and medium-scale renewable installations. This paper introduces a blockchain-based carbon-credit certification architecture, demonstrated through a 100 kWp photovoltaic case study, that integrates real-time IoT data collection, edge-level aggregation, and secure on-chain storage on a permissioned blockchain with smart contracts. Unlike approaches focused on trading mechanisms, the proposed system aligns with European legislation and voluntary carbon-market standards, clarifying the practical requirements and constraints that apply to photovoltaic operators. The resulting architecture provides a structured pathway for generating verifiable carbon-credit records and supporting third-party verification.

Aiming at the core problem that it is difficult for a fixed inertia coefficient to balance transient disturbance suppression and long-term stability in complex network synchronization systems, an adaptive inertia control strategy based on variational optimization is proposed. Taking the Kuramoto model with inertia as the research carrier, the analytical expression of the time-varying inertia coefficient M(t) is strictly derived by the functional variational method, and a hierarchical control structure of "benchmark inertia + disturbance feedback" is constructed to achieve the organic unity of minimizing the vulnerability performance function H(T) and stability constraints. A multimodal decoupling control strategy based on Laplacian eigenvector projection is designed to enhance the feedback strength of the dominant mode by eigenvalue weighting, improving the control accuracy and dynamic response speed. Simulation verification is carried out in complex network systems, and the control performance of regular networks (RG), random networks (ER), small-world networks (SW), scale-free networks (SF) and spider webs (SP) under three typical disturbances of pulses, monotonic decays and oscillatory decays is systematically analyzed. The results show that the proposed strategy reduces H(T) of the five networks by 19%-25%, shortens the relaxation time by 15%-24%, and the real parts of all system eigenvalues are less than -0.25s^-1 , meeting the asymptotic stability criterion. This study provides a new theoretical framework and engineering implementation scheme for the stability control of complex network synchronization systems, which can be widely applied to fields such as power grids, communication networks, and neural networks.

This paper presents the flexible RIM Hand, a biomimetic robotic hand that precisely replicates the carpometacarpal (CMC) joints and employs superelastic Nitinol wires throughout its skeletal framework. By modeling the full carpal-to-metacarpal anatomy, the design enables realistic palm deformation through tendon-driven fingers while enhancing joint restoration and supports skeletal structure with Nitinol-based dorsal extensors. A flexible silicone skin further increases contact friction and contact area, enabling stable grasps for diverse objects. Experiments show that the palm can deform up to 28%, matching human hand flexibility, while achieving more than twice the payload capacity and three times the contact area compared to a rigid palm design. The RIM Hand thus offers improved dexterity, compliance, and anthropomorphism, making it promising for prosthetic and service-robot applications.

In speech machine learning, neural network models are typically designed by choosing an architecture with fixed layer sizes and structure. These models are then trained to maximize performance on metrics aligned with the task's objective. While the overall architecture is usually guided by prior knowledge of the task, the sizes of individual layers are often chosen heuristically. However, this approach does not guarantee an optimal trade-off between performance and computational complexity; consequently, post hoc methods such as weight quantization or model pruning are typically employed to reduce computational cost. This occurs because stochastic gradient descent (SGD) methods can only optimize differentiable functions, while factors influencing computational complexity, such as layer sizes and floating-point operations per second (FLOP/s), are non-differentiable and require modifying the model structure during training. We propose a reparameterization technique based on feature noise injection that enables joint optimization of performance and computational complexity during training using SGD-based methods. Unlike traditional pruning methods, our approach allows the model size to be dynamically optimized for a target performance-complexity trade-off, without relying on heuristic criteria to select which weights or structures to remove. We demonstrate the effectiveness of our method through three case studies, including a synthetic example and two practical real-world applications: voice activity detection and audio anti-spoofing. The code related to our work is publicly available to encourage further research.

The inherent non-convexity of poriferous surfaces typically entraps agents in local minima and complicates workload distribution. To resolve this, we propose a distributed diffeomorphic coverage control framework for the multi-agent system (MAS) in such surfaces. First, we establish a distributed poly-annulus conformal mapping that transforms arbitrary poriferous surfaces into a multi-hole disk. Leveraging this topological equivalence, a collision-free sectorial partition mechanism is designed in the multi-hole disk, which rigorously induces strictly connected subregions and workload balance on the poriferous surfaces. This mechanism utilizes a buffer-based sequence mechanism to ensure strict topological safety when bypassing obstacles. Furthermore, a pull-back Riemannian metric is constructed to define the length metric that encodes safety constraints. Based on this metric, a distributed gradient-based control law is synthesized to drive agents toward optimal configurations, ensuring simultaneous obstacle avoidance and coverage optimization. Theoretical analyses guarantee the Input-to-State Stability (ISS) of the partition dynamics and the asymptotic convergence of the closed-loop system. Numerical simulations confirm the reachability and robustness of the proposed coverage algorithm, offering a scalable solution for distributed coverage in poriferous surfaces.

Agentic artificial intelligence (AI) is an AI paradigm that can perceive the environment, reason over observations, and execute actions to achieve specific goals. Task-oriented communication supports agentic AI by transmitting only the task-related information instead of full raw data in order to reduce the bandwidth requirement. In real-world scenarios, AI agents often need to perform a sequence of actions to complete complex tasks. Completing these long-horizon tasks requires a hierarchical agentic AI architecture, where a high-level planner module decomposes a task into subtasks, and a low-level actor module executes each subtask sequentially. Since each subtask has a distinct goal, the existing task-oriented communication schemes are not designed to handle different goals for different subtasks. To address this challenge, in this paper, we develop a hierarchical task-oriented communication (HiTOC) framework. We consider a system with an edge server and a robot as an edge device. The high-level planner and low-level actor modules reside on the edge server. The robot transmits only the environmental observation that is relevant to the current subtask to the edge server. We propose a conditional variational information bottleneck (cVIB) approach to train the HiTOC framework to adaptively transmit minimal information required for each subtask. Simulations conducted on the AI2-THOR platform demonstrate that the proposed HiTOC framework outperforms three state-of-the-art schemes in terms of the success rate on MAP-THOR benchmark.

With the growing applications of the Internet of Things (IoT), a major challenge is to ensure continuous connectivity while providing prioritized access. In dense IoT scenarios, synchronization may be disrupted either by the movement of nodes away from base stations or by the unavailability of reliable Global Navigation Satellite System (GNSS) signals, which can be affected by physical obstructions, multipath fading, or environmental interference, such as such as walls, buildings, moving objects, or electromagnetic noise from surrounding devices. In such contexts, distributed synchronization through Non-Orthogonal Multiple Access (NOMA) offers a promising solution, as it enables simultaneous transmission to multiple users with different power levels, supporting efficient synchronization while minimizing the signaling overhead. Moreover, NOMA also plays a vital role for dynamic priority management in dense and heterogeneous IoT environments. In this article, we proposed a Two-Stage NOMA-Enabled Framework "TSN-IoT" that integrates the mechanisms of conventional Precision Time Protocol (PTP) based synchronization, distributed synchronization and data transmission. The framework is designed as a four-tier architecture that facilitates prioritized data delivery from sensor nodes to the central base station. We demonstrated the performance of "TSN-IoT" through a healthcare use case, where intermittent connectivity and varying data priority levels present key challenges for reliable communication. Synchronization speed and end-to-end delay were evaluated through a series of simulations implemented in Python. Results show that, compared to priority-based Orthogonal Frequency Division Multiple Access (OFDMA), TSN-IoT achieves significantly better performance by offering improved synchronization opportunities and enabling parallel transmissions over the same sub-carrier.

Compressed sensing, which involves the reconstruction of sparse signals from an under-determined linear system, has been recently used to solve problems in group testing. In a public health context, group testing aims to determine the health status values of p subjects from n<<p pooled tests, where a pool is defined as a mixture of small, equal-volume portions of the samples of a subset of subjects. This approach saves on the number of tests administered in pandemics or other resource-constrained scenarios. In practical group testing in time-constrained situations, a technician can inadvertently make a small number of errors during pool preparation, which leads to errors in the pooling matrix, which we term `model mismatch errors' (MMEs). This poses difficulties while determining health status values of the participating subjects from the results on n<<p pooled tests. In this paper, we present an algorithm to correct the MMEs in the pooled tests directly from the pooled results and the available (inaccurate) pooling matrix. Our approach then reconstructs the signal vector from the corrected pooling matrix, in order to determine the health status of the subjects. We further provide theoretical guarantees for the correction of the MMEs and the reconstruction error from the corrected pooling matrix. We also provide several supporting numerical results.

Orthogonal time frequency space (OTFS) modulation stands out as a promising waveform for sixth generation (6G) and beyond wireless communication systems, offering superior performance over conventional methods, particularly in high-mobility scenarios and dispersive channel conditions. Recent research has demonstrated that the reduced computational complexity of deep learning (DL)-based signal detection (SD) methods constitutes a compelling alternative to conventional techniques. In this study, low-complexity DL-based SD methods are proposed for a multiple-input multiple-output (MIMO)-OTFS system and examined under Nakagami-$m$ channel conditions. The symbols obtained from the receiver antennas are combined using maximum ratio combining (MRC) and detected with the help of a DL-based detector implemented with multi-layer perceptron (MLP), convolutional neural network (CNN), and residual network (ResNet). Complexity analysis reveals that the MLP architecture offers significantly lower computational complexity compared to CNN, ResNet, and classical methods such as maximum likelihood detection (MLD). Furthermore, numerical analyses have shown that the proposed DL-based detectors, despite their low complexity, achieve comparable bit error rate (BER) performance to that of a high-performance MLD under various system conditions.

We present S$^2$Voice, the winning system of the Singing Voice Conversion Challenge (SVCC) 2025 for both the in-domain and zero-shot singing style conversion tracks. Built on the strong two-stage Vevo baseline, S$^2$Voice advances style control and robustness through several contributions. First, we integrate style embeddings into the autoregressive large language model (AR LLM) via a FiLM-style layer-norm conditioning and a style-aware cross-attention for enhanced fine-grained style modeling. Second, we introduce a global speaker embedding into the flow-matching transformer to improve timbre similarity. Third, we curate a large, high-quality singing corpus via an automated pipeline for web harvesting, vocal separation, and transcript refinement. Finally, we employ a multi-stage training strategy combining supervised fine-tuning (SFT) and direct preference optimization (DPO). Subjective listening tests confirm our system's superior performance: leading in style similarity and singer similarity for Task 1, and across naturalness, style similarity, and singer similarity for Task 2. Ablation studies demonstrate the effectiveness of our contributions in enhancing style fidelity, timbre preservation, and generalization. Audio samples are available~\footnote{https://honee-w.github.io/SVC-Challenge-Demo/}.

Hybrid microgrids integrating Grid-Following (GFL) and Grid-Forming (GFM) inverters present complex control challenges arising from the decoupling between long-term economic dispatch and real-time dynamic regulation, as well as the distinct physical limitations of heterogeneous inverters under cyber uncertainties. This paper proposes a Resilient Hierarchical Power Control (RHPC) strategy to unify these conflicting requirements within a cohesive framework. A standardized power increment mechanism is developed to bridge the tertiary and secondary layers, ensuring that real-time load fluctuations are compensated strictly according to the optimal economic ratios derived from the tertiary layer. To address the strict active power saturation constraints of GFL units, a dynamic activation scheme coupled with projection operators is introduced, which actively isolates saturated nodes from the consensus loop to prevent integrator wind-up and preserve the stability of the GFM backbone. Furthermore, the proposed framework incorporates a multi-scale attention mechanism and LSTM-based predictors into the secondary control protocol, endowing the system with robustness against unbounded False Data Injection (FDI) attacks and packet losses. Rigorous theoretical analysis confirms that the system achieves Uniformly Ultimately Bounded (UUB) convergence, and simulations on a modified IEEE 33-bus system demonstrate that the proposed strategy significantly improves power sharing accuracy and operational resilience in both grid-connected and islanded modes compared to conventional methods.

Power system outages expose market participants to significant financial risk unless promptly detected and hedged. We develop an outage identification method from public market signals grounded in the parametric quickest change detection (QCD) theory. Parametric QCD operates on stochastic data streams, distinguishing pre- and post-change regimes using the ratio of their respective probability density functions. To derive the density functions for normal and post-outage market signals, we exploit multi-parametric programming to decompose complex market signals into parametric random variables with a known density. These densities are then used to construct a QCD-based statistic that triggers an alarm as soon as the statistic exceeds an appropriate threshold. Numerical experiments on a stylized PJM testbed demonstrate rapid line outage identification from public streams of electricity demand and price data.

This study presents a real-world canine heart sound dataset and evaluates SoNUS version 3.2.x, a machine learning algorithm for preliminary cardiac analysis using smartphone microphone recordings. More than one hundred recordings were collected from dogs across four continents, with thirty eight recordings annotated by board certified veterinary cardiologists for quantitative evaluation. SoNUS version 3.2.x employs a multi-stage fallback architecture with quality-aware filtering to ensure reliable output under variable recording conditions. The primary sixty second model achieved mean and median heart rate accuracies of ninety one point six three percent and ninety four point nine five percent, while a fast model optimized for thirty to forty second recordings achieved mean and median accuracies of eighty eight point eight six percent and ninety two point nine eight percent. These results demonstrate the feasibility of extracting clinically relevant cardiac information from opportunistic smartphone recordings, supporting scalable preliminary assessment and telehealth applications in veterinary cardiology.

We study a two-type server queueing system where flexible Type-I servers, upon their initial interaction with jobs, decide in real time whether to process them independently or in collaboration with dedicated Type-II servers. Independent processing begins immediately, as does collaborative service if a Type-II server is available. Otherwise, the job and its paired Type-I server wait in queue for collaboration. Type-I servers are non-preemptive and cannot engage with new jobs until their current job is completed. We provide a complete characterization of the structural properties of the optimal policy for the clearing system. In particular, an optimal control is shown to follow a threshold structure based on the number of jobs in the queue before a Type-I first interaction and on the number of jobs in either independent or collaborative service. We propose simple threshold heuristics, based on linear approximations, for real-time decision-making. In much of the parameter and state spaces, we establish theoretical bounds that compare the thresholds proposed by our heuristics to those of optimal policies and identify parameter configurations where these bounds are attained. Outside of these regions, the optimal thresholds are infinite. Numerical experiments further demonstrate the accuracy and robustness of our heuristics, particularly when the initial queue length is high. Our proposed heuristics achieve costs within 0.5% of the optimal policy on average and significantly outperform benchmark policies that exhibit extreme sensitivity to system parameters, sometimes incurring costs exceeding 100% of the optimal.

We study a two-stage tandem service queue attended by two servers. Each job-server pair must complete both service phases together, with the server unable to begin a new job until the current one is fully processed after two stages. Immediately after the first phase of service, the server decides whether to send the job/customer to a downstream station that allows parallel processing or to a single-service facility that offers faster or higher-quality service but handles only one job at a time. This choice determines whether the second phase commences immediately or (potentially) after waiting in a queue for the single-service facility to become available. The decision-making scenario is modeled via a Markov decision process formulation, of a clearing system with holding costs at each station. We fully characterize the structural properties of an optimal control policy based on the relationship between the service rates at the downstream stations. A numerical study highlights the significance of optimal control by comparing its performance against several natural heuristic policies.

Open-vocabulary 6D object pose estimation empowers robots to manipulate arbitrary unseen objects guided solely by natural language. However, a critical limitation of existing approaches is their reliance on unconstrained global matching strategies. In open-world scenarios, trying to match anchor features against the entire query image space introduces excessive ambiguity, as target features are easily confused with background distractors. To resolve this, we propose Fine-grained Correspondence Pose Estimation (FiCoP), a framework that transitions from noise-prone global matching to spatially-constrained patch-level correspondence. To systematically eliminate background interference, FiCoP first employs an object-centric disentanglement step to isolate the target from macro-level environmental noise. Building upon this localized region, our core methodological innovations are twofold. Firstly, a Cross-Perspective Global Perception (CPGP) module is proposed to fuse dual-view features, establishing structural consensus through explicit context reasoning and text-guided semantic injection. Secondly, we design a Patch Correlation Predictor (PCP) that leverages a patch-to-patch correlation matrix as a structural prior. This generates a precise block-wise association map, acting as a spatial filter to enforce fine-grained, noise-resilient matching. Experiments on the REAL275 and Toyota-Light datasets demonstrate that FiCoP improves Average Recall by 8.0% and 6.1%, respectively, compared to the state-of-the-art method, highlighting its capability to deliver robust and generalized perception for robotic agents operating in complex, unconstrained open-world environments. The source code will be made publicly available at https://github.com/zjjqinyu/FiCoP.

In this paper, we study robust beamforming design for near-field physical-layer-security (PLS) systems, where a base station (BS) equipped with an extremely large-scale array (XL-array) serves multiple near-field legitimate users (Bobs) in the presence of multiple near-field eavesdroppers (Eves). Unlike existing works that mostly assume perfect channel state information (CSI) or location information of Eves, we consider a more practical and challenging scenario, where the locations of Bobs are perfectly known, while only imperfect location information of Eves is available at the BS. We first formulate a robust optimization problem to maximize the sum-rate of Bobs while guaranteeing a worst-case limit on the eavesdropping rate under location uncertainty. By transforming Cartesian position errors into the polar domain, we reveal an important near-field angular-error amplification effect: for the same location error, the closer the Eve, the larger the angle error, severely degrading the performance of conventional robust beamforming methods based on imperfect channel state information. To address this issue, we first establish the conditions for which the first-order Taylor approximation of the near-field channel steering vector under location uncertainty is largely accurate. Then, we propose a two-stage robust beamforming method, which first partitions the uncertainty region into multiple fan-shaped sub-regions, followed by the second stage to formulate and solve a refined linear-matrix-inequality (LMI)-based robust beamforming optimization problem. In addition, the proposed method is further extended to scenarios with multiple Bobs and multiple Eves. Finally, numerical results validate that the proposed method achieves a superior trade-off between rate performance and secrecy robustness, hence significantly outperforming existing benchmarks under Eve location uncertainty.

The ICASSP 2026 URGENT Challenge advances the series by focusing on universal speech enhancement (SE) systems that handle diverse distortions, domains, and input conditions. This overview paper details the challenge's motivation, task definitions, datasets, baseline systems, evaluation protocols, and results. The challenge is divided into two complementary tracks. Track 1 focuses on universal speech enhancement, while Track 2 introduces speech quality assessment for enhanced speech. The challenge attracted over 80 team registrations, with 29 submitting valid entries, demonstrating significant community interest in robust SE technologies.