Vehicle body type is a significant determinant of cyclist injury severity in overtaking crashes, yet automated tools for classifying vehicles into injury-risk-relevant categories from naturalistic roadway video do not exist in the open literature. Standard object detection benchmarks provide only coarse vehicle labels (car, truck, bus, motorcycle), while existing fine-grained recognition systems are trained on controlled imagery and lack evaluation for deployment robustness across recording sites. This paper presents an open-source two-stage computer vision pipeline combining a pre-trained RT-DETR detector for coarse vehicle localization with a fine-tuned Vision Transformer (ViT-Base/16) for six-category body-type classification: passenger car, SUV, pickup truck, minivan, large van, and commercial truck. A confidence-based abstention mechanism withholds Stage 2 predictions when softmax output falls below 0.60, producing unknown labels rather than silent misclassifications. Evaluated on 3,805 annotated overtaking events from a bicycle-lane corridor in Ann Arbor, Michigan (in-distribution), the pipeline achieved 0.94 accuracy with per-class F1 scores from 0.91 (minivan) to 0.97 (SUV). On an independent out-of-distribution evaluation of 311 events from an open cycling dataset without retraining, accuracy was 0.89. Three of four well-represented categories maintained F1 at or above 0.90 under domain shift. The largest degradation was observed for minivan (F1 = 0.72), driven by abstention rate rising from 2.4% to 25.0% rather than active misclassification, consistent with the mechanism propagating genuine model uncertainty. The full pipeline, including inference scripts, training code, evaluation utilities, and model weights, is released as open-source software to support reproducibility and reuse across roadside video archives and cycling safety research.

Audio is an inherently interactive modality, yet today's Large Audio Language Models (LALMs) are offline, and streaming audio models each handle only a single task such as streaming ASR or voice chatting. It is time to unify them into one online LALM: a model that, through an always-on perceive-decide-respond loop, listens to sound, environment, and instructions in real time and reacts on the fly. We formalize this regime as the Audio Interaction Model, and realize it with Audio-Interaction, a unified streaming model that retains offline task execution while adding online general audio instruction following, from dialogue to full voice chatting, deciding when to respond from the semantics of the stream. To enable this, we propose SoundFlow, a framework that instantiates the perceive-decide-respond loop end to end, from data to training to deployment, through streaming-native data construction, comprehension-aware training, and asynchronous low-latency inference for stable real-time interaction. We further construct StreamAudio-2M, a 2.6M-item streaming corpus spanning 7 fundamental abilities and 28 sub-tasks, and Proactive-Sound-Bench for evaluating proactive audio intervention. Across 8 benchmarks, Audio-Interaction preserves competitive performance on mainstream audio tasks while unlocking capabilities inaccessible to offline LALMs, including real-time ASR, streaming audio instruction following, and proactive help.

Radiotherapy is a cornerstone of glioma treatment inducing complex structural changes in brain tissue that are difficult to anticipate. Predicting these changes from pretreatment data could improve understanding of treatment-related effects and support the development of image-based outcome prediction methods. Recent studies have shown that follow-up brain magnetic resonance imaging can be synthesized from baseline imaging and treatment information, but most existing approaches operate on single 2D slices and represent treatment as a global parameter, rather than a spatially dynamic variable. In this work, we address both limitations with a 3D latent diffusion framework that conditions image generation on the spatially resolved voxel-wise dose distribution, alongside a pretreatment image and follow-up time. To make volumetric synthesis computationally feasible, the model combines latent-space compression with ControlNet-based spatial conditioning. The method was trained and evaluated on a public dataset comprising 257 scans from 25 glioma patients. Prediction quality was assessed using mean squared error, peak signal-to-noise ratio, and structural similarity index. Anatomical consistency was further evaluated using Dice scores for cerebrospinal fluid, gray matter, and white matter segmentations, together with hippocampus volume prediction error and deformation analysis based on log Jacobian determinant maps. Compared with our previously proposed 2D approach, the 3D model achieved improved image similarity while maintaining good agreement with ground truth anatomy and deformation patterns. Overall, these results support the feasibility of 3D treatment-aware generative modeling for predicting post-radiotherapy brain MRI using only pretreatment information. Code is available at https://github.com/nordinbelkacemi/fu-pred-3d

We consider radar sensing with affine frequency division multiplexing (AFDM), a chirp-based waveform recently proposed for high-mobility integrated sensing and communication. While numerical Cramér-Rao bounds for AFDM radar are available in the literature, no closed-form Fisher information analysis has so far revealed how the waveform's chirp structure shapes delay-Doppler estimation accuracy.In this paper, we provide such an analysis. We identify a cancellation in the AFDM likelihood: the frequency drift introduced by the chirp modulation is exactly compensated by a discrete phase correction built into the chirp-periodic prefix, leaving only a small residual. Exploiting this cancellation, we derive an exact closed-form Fisher information matrix that depends on the AFDM chirp structure through a single scalar, and from it we obtain closed-form Cramér-Rao bounds for joint delay and Doppler estimation.Three consequences follow. AFDM is provably less delay-Doppler-coupled than OFDM for any nonzero chirp rate. The delay Cramér-Rao bound improves quadratically with the chirp rate, while the Doppler bound is unaffected by it. Finally, our framework reduces continuously to the classical OFDM result as the chirp vanishes, certifying it as a strict generalization of OFDM radar sensing theory.Overall, our work shows that the chirp-periodic prefix -- until now studied only as a channel-equalization device -- is the structural element that decouples delay and Doppler in AFDM sensing, and that AFDM's superior sensing performance can be characterized analytically rather than through numerical bounds alone. Numerical experiments at realistic vehicular and low-Earth-orbit parameters validate all closed-form expressions.

We develop an adaptive OFDM framework for underwater acoustic communications based on PatchCSI-T, a Transformer-based multistep channel prediction model with feature-independent modeling and parameter sharing. Combined with a greedy adaptive modulation and power allocation scheme, the proposed approach enables accurate, low-latency CSI forecasting and improves end-to-end BER and spectral efficiency on real-world UWA channel datasets.

A compact dynamic four-element array with omnidirectional H-plane coverage is presented for planar physical-layer security using antenna-level directional modulation. The proposed approach achieves angularly selective information transmission without phased-array beamforming or multiple RF chains by dynamically switching the excitation paths of a four-element array. The antenna comprises four printed meander-line monopole elements operating at 5.05 GHz with independently controlled differential power excitation, which introduces magnitude and phase pattern modulation and dynamic motion of the apparent element spacing, resulting in strongly angle-dependent signal distortion and bit error rate (BER) performance. Reliable information recovery is confined to a narrow broadside region in the E-plane, while significantly elevated BER is observed at off-broadside angles. In contrast, the H-plane radiation remains static and omnidirectional, enabling full 360-degree information-recoverable coverage in the orthogonal plane. The antenna is fabricated on a single-layer Rogers RO4350B substrate with a compact footprint of 0.55 x 1.73 lambda_0^2. A four-path switching network implemented using commercial RF components validates the concept experimentally. Communication measurements under high-SNR conditions above 19 dB using 16-QAM demonstrate a planar information beamwidth below 24 degrees, confirming effective antenna-level directional modulation with angle-dependent BER characteristics and omnidirectional H-plane coverage.

This paper proposes a novel approach for the synchronization problem of nonlinear dynamical systems, integrating dual Lyapunov stability analysis with polynomial optimization. A comprehensive review of the relevant scientific literature on synchronization methods is conducted, with a particular focus on classical Lyapunov-based methods for chaotic systems. In this study, the Rössler system is synchronized by employing dual Lyapunov-based closed-loop synchronization method. This method uses semidefinite programming and sum-of-squares polynomials to compute a nonlinear state feedback function which synchronize a chaotic system to a selected reference model. It is aimed that chaotic behavior is destroyed and, instead, a limit cycle becomes attracting. Simulation works are performed for randomly selected 100 different initial conditions to show that synchronization process is successfully performed. Furthermore, bifurcation diagrams and phase portraits are evaluated to analyze the system dynamics. The paper discusses results and how new constraints should be employed and adapted to more complex systems.

Energy-intensive data centers (DCs) have emerged as substantial and flexible loads in modern power systems, underscoring the critical need for computation-electricity coordination. Harnessing the spatio-temporal flexibility of DC workloads is a promising approach to facilitate this coordination. However, existing studies overlook the collaborative potential of computational resource sharing among geo-distributed DCs, thereby failing to fully unlock this flexibility. In this paper, a bi-level computation-electricity coordination framework is proposed to explicitly capture the bidirectional interactions between DCs and power grid. Firstly, a peer-to-peer cloud service market (P2P-CSM) for geo-distributed DCs is proposed, which enables bilateral cloud service transactions to leverage regional heterogeneities (e.g., electricity prices, cooling efficiency). Secondly, locational marginal prices are embedded into the framework to reflect network congestion and nodal price disparities. Thirdly, a dual consensus alternating direction method of multipliers (ADMM)-based decentralized algorithm is developed as the P2P market clearing algorithm, and a bisection-assisted iterative algorithm is proposed to ensure rigorous convergence of the framework. Case studies conducted on modified IEEE 30-bus system validate that the P2P-CSM achieves a win-win computation-electricity coordination: it not only increases total DC operational profit by 22.8\%, but also effectively alleviates grid congestion and yields a 3.2\% reduction in total energy consumption.

The growing complexity of modern energy systems has led to the adoption of Smart Grid (SG) that use advanced communication technologies to facilitate efficient, reliable, secure, and sustainable energy operation and management. Unlike existing surveys that often treat grid and communication domains separately, this work rigorously quantifies service requirements for high-complexity emerging scenarios. It provides a comprehensive overview of SG architecture that integrates digital communication infrastructure with distributed energy resources (DERs), microgrids, energy storage systems, and cybersecurity frameworks. Furthermore, emerging SG use cases such as smart distributed voltage control, real-time fault detection and self-healing, smart and autonomous monitoring, and predictive maintenance are identified, and more importantly, service performance requirements associated with these use cases have been quantified. Additionally, key capabilities and emerging SG enablers of fifth-generation (5G) and sixth-generation (6G) networks are described. These capabilities and enablers include network slicing, edge computing, spectrum management, artificial intelligence (AI) driven optimization, digital twins, and Open-Radio Access Network (O-RAN). Finally, the paper discusses open challenges and future research directions for designing scalable, intelligent, and secure next-generation SG systems.

Sygyt is a Tuvan style of biphonic singing in which a low vocal drone is sustained while a high harmonic is selectively amplified in the 1--3\,kHz region. Copy-synthesizing this effect remains challenging for articulatory models, since it requires fine control of narrowly focused resonances that standard low-dimensional tract parameterizations cannot easily reproduce. We address this problem with a differentiable Kelly--Lochbaum waveguide augmented with a sublingual second source, cubic B-spline tract parameterization, and spatially varying learnable damping, optimized end-to-end by gradient descent from audio. On 20 segments from two independent sygyt datasets (5 singers, 10 pitches), the proposed model reduces log-spectral distance by 30--38\% relative to an articulatory baseline, with the largest gains concentrated in the overtone region. Cepstral-envelope analysis further shows more accurate recovery of the merged formant structure characteristic of sygyt production. The model also outperforms a DDSP harmonic-plus-noise baseline with direct per-harmonic spectral control, suggesting that explicit acoustic structure is a useful inductive bias for overtone-singing copy-synthesis.

The escalating accumulation of orbital debris poses a critical threat to future space operations. Space-based lasers leveraging laser ablation have emerged as a promising approach for mitigating debris proliferation and preserving the orbital environment. Current literature, however, treats space-based laser debris remediation as a deterministic problem, assuming that momentum transfer and the resulting debris perturbations are precisely known. In reality, laser-to-debris engagement outcomes are inherently stochastic due to partially known debris characteristics. Compounding this challenge, estimating critical laser-matter parameters in situ, such as the momentum coupling coefficient, requires ablation that consequently perturbs the debris trajectory. This establishes a coupled ablation-and-estimation problem in which the laser platform and target debris encounter geometry influences remediation effectiveness and estimation accuracy. To address this problem, we present a joint ablation-and-estimation methodology that provides insights into the driving factors that make different encounter geometries improve or degrade overall remediation and estimation performance. Results across multiple coplanar and out-of-plane encounter geometries demonstrate how periapsis-lowering capacity, linear system observability, and nonlinear estimation performance evolve as laser parameters and relative orbit geometry vary. By identifying the key drivers behind these metrics, this study highlights critical considerations for the safe and effective operation of space-based lasers under uncertainty.

Audio generation and audio-to-text understanding remain largely separate, with diffusion models dominating high-fidelity synthesis and autoregressive (AR) language models driving captioning and semantic prediction. Existing unified approaches typically rely on either heterogeneous modules or AR-centric modeling, which can hinder joint optimization and limit acoustic fidelity. We present UAT, to our knowledge, the first diffusion-centric framework that supports unified audio generation, editing, and captioning. UAT couples continuous latent diffusion for audio with masked discrete diffusion for text, enabling bidirectional audio-text modeling within a shared dual-stream backbone. Experiments show that UAT preserves strong audio generation and editing capabilities while achieving competitive captioning performance, demonstrating a favorable balance between acoustic synthesis and semantic prediction. Demo samples are available at https://UAT-demo.github.io.

The goal of single-channel source separation is to reconstruct $K$ sources given their mixture. In supervised settings where vast amounts of clean source data are available, this challenging, ill-posed problem has been addressed successfully by generative diffusion and flow-based prior models. However, access to such clean source samples is often limited, and even when available, supervised models are vulnerable to domain shifts. To bridge this gap, we present Separation via Unsupervised Remixing Flow (SURF), an unsupervised flow matching approach for source separation that learns directly from observed mixtures. This method relies on a novel combination of state-of-the-art supervised flow matching and regression-based self-supervised techniques. At a high level, starting from a teacher model, we utilize a "remixing" step to bootstrap the learning of a student flow model from the teacher's estimates. We provide insights into the objectives optimized by this approach and draw a novel connection to the Wake-Sleep algorithm. Empirical evaluations on image and audio benchmarks demonstrate that SURF establishes a new state-of-the-art, significantly outperforming existing unsupervised methods. See our demo page for examples. https://google.github.io/df-conformer/surf/

This paper proposes an access protocol framework for segmented waveguide-enabled pinching-antenna systems (SWANs), which exploits SWAN-induced reconfigurable channel diversity as a protocol-level resource for uplink random access. The framework consists of two stages, a channel-oracle stage and an access stage, designed under three SWAN operating modes: (i) one-segment selection (OS), (ii) segment aggregation (SA), and (iii) segment multiplexing (SM). Specifically, in the channel oracle stage, the OS mode is adopted to acquire sparse pilot observations and infer the channel responses across the SWAN configuration space. In this way, high-dimensional uplink channel acquisition is recast as a low-dimensional geometric localization problem, thereby reducing pilot overhead while preserving channel reconstruction accuracy. For the access stage, we construct two oracle-guided access codebooks under the SA and SM modes, respectively, which address the tradeoff between hardware complexity and multiuser access resolution. In particular, the SA-based scheme supports single radio frequency (RF) chain access through randomized segment-group activation, whereas the SM-based R-access scheme exploits multiple RF chains to construct deterministic access slots and enhance collision resolution. Finally, our numerical results demonstrate that (i) the proposed two-stage framework improves access performance under the same training overhead, (ii) anchor densification is more effective than aggressive segment aggregation for SA, and (iii) SM-based R-access achieves deterministic coverage and higher throughput in moderate- and high-load regimes, whereas SA-based access remains attractive for low-complexity implementations.

Cooperative localization (CL) is fundamental in emerging multi-agent systems, where agents fuse local sensing data with exchanged information to estimate their own states. At a large scale, however, tracking cross-correlations becomes infeasible, preventing the use of optimal filters. Ignoring or underestimating these correlations leads to overconfident, and thus inconsistent, estimates. Existing CL algorithms achieve good performance and consistency typically at the expense of communication, computation, or memory that scales with the network size. This is incompatible with ultra large-scale systems (ULSS) - for example, satellite mega-constellations - where per-agent resources are limited and must remain independent of the number of agents. This reveals a critical gap: no existing CL method is simultaneously well-performing, consistent, and ULSS-scalable. This paper introduces a new CL framework that addresses this gap using the recently proposed overlapping covariance intersection methodology, which enables agents to exploit limited structural information about cross-correlations without compromising consistency. The resulting CL algorithm leads to optimal conservative covariance propagation using only locally available information. The method is fully distributed, scalable to an ultra large scale, and provably recursively consistent. Simulations demonstrate substantial performance improvement over state-of-the-art consistent CL approaches while preserving scalability.

The rapid growth of hyperscale AI datacenters introduces structured, workload-driven active-power fluctuations at the point of interconnection. These fluctuations appear to the grid as time-varying disturbance injections that cannot be captured by conventional peak- or average-load representations. To reduce the residual power disturbance before it propagates into the bulk power system, this paper proposes a hybrid energy storage system with differentiable predictive control (HESS-DPC) framework for datacenter-side power smoothing. A workload-driven disturbance model is first established, representing the point-of-interconnection load deviation as the superposition of training and fine-tuning workloads to capture the structured forcing inputs that can excite generator frequency dynamics. A frequency-based rule-based controller then allocates this deviation between a battery energy storage system (BESS) and a supercapacitor (SC), assigning the energy-dominant component to the BESS and the fast-varying component to the SC. To overcome the anticipation and constraint limitations of fixed-frequency decomposition, a residual differentiable predictive control policy is trained offline to compute finite-horizon command corrections around the rule-based baseline while enforcing a one-step safeguard. Simulations on the NPCC 140-bus system show that HESS-DPC reduces grid-side residual deviations during workload transitions, improves SC state-of-charge sustainability over extended operation, and reduces generator peak-to-peak frequency deviations by more than 80 percent across all monitored generators, with the worst-affected generator response falling from 15.1 mHz to 1.3 mHz. These results confirm that local active-power smoothing at the datacenter point of interconnection can substantially mitigate frequency disturbances caused by AI workloads.

Future 6G networks are envisioned to integrate low Earth orbit satellite mega-constellations to enable seamless global connectivity, particularly in underserved and remote areas. However, the deployment of dense mega-constellations introduces interference among satellites operating over shared frequency bands. This represents a rather new setup for studying spectrum sharing, which exacerbates the limited flexibility of conventional FDD systems based on fixed bands for downlink and uplink transmissions. We address this spectrum-sharing problem and propose dynamic re-assignment of FDD bands for improved interference management in dense deployments, as well as evaluate the performance gain of this approach. To this end, we formulate a joint optimization problem that incorporates dynamic band assignment, user scheduling, and power allocation in both directions. This non-convex mixed integer problem is solved using a combination of equivalence transforms, alternating optimization, and state-of-the-art industrial-grade mixed integer solvers. Numerical results demonstrate that the proposed approach of dynamic FDD band assignment significantly enhances system performance over conventional FDD, achieving up to 30\% improvement in throughput in dense deployments.

Control barrier functions (CBFs) are a widely applied modular tool to ensure safe operation of nonlinear dynamical control systems. However, for their construction accurate knowledge of the system dynamics is typically needed. This requirement was recently alleviated for relative-degree-one systems using techniques from prescribed performance control (PPC) or funnel control (FC). This article extends the model-free CBF design to nonlinear systems of arbitrary relative degree. Moreover, we show with a simple example that a straightforward extension of existing results for relative-degree-one systems fails. Instead, we utilize novel techniques from funnel control to characterize a subset of the controls satisfying a CBF condition without requiring a dynamic model or state measurement. Finally, we demonstrate the applicability of our results on a seven degrees of freedom robotic manipulator with relative degree two.

Despite the widespread use of PID controllers in engineering practice, designing optimal PID parameters has long been regarded as a challenging problem in both theory and practice, particularly when faced with uncertain nonlinear dynamical systems. Based on the authors' PID control theory established recently for MIMO nonlinear uncertain systems (Zhao and Guo, 2022), which provides a concrete PID parameter set for global stability of PID controlled systems, this paper further proposes a near-optimal PID tuning method, where only input-output (zeroth-order) data on the control performance is available. The tuning method is formulated as a constrained optimization problem and solved by an iterative learning algorithm, referred to as HRS-KW algorithm, that combines a hysteretic random search with the Kiefer-Wolfowitz algorithm, aiming at utilizing the advantages of both global exploration and local gradient acceleration. This method operates without requiring precise structural knowledge of the system dynamics, yet its almost sure convergence to an epsilon-optimal solution for the PID parameters can be guaranteed in theory while ensuring closed-loop system stability. Simulation results illustrate that our HRS-KW algorithm outperforms other related optimization methods, exhibiting better convergence to the prescribed epsilon-optimal performance set.

Text-to-video (T2V) models trained on large-scale web data can generate undesired content, motivating interventions that reduce harmful outputs without sacrificing visual quality. Activation steering offers an attractive mechanistic alternative to finetuning and prompt filtering, but existing T2V steering methods remain limited, typically applying coarse, non-anticipative interventions that can lead to oversteering and content degradation. To close this gap, we propose Latent Activation Linear-Quadratic Regulator (LA-LQR), a reduced-order optimal control framework for minimally invasive T2V steering. LA-LQR formulates T2V inference as a dynamical system and computes closed-loop feedback interventions that steer activations toward desired feature setpoints while penalizing unnecessary perturbations. To make optimal control feasible for high-dimensional video activations, we project activations onto a low-dimensional, task-relevant subspace derived from contrastive prompt pairs, estimate local linear dynamics in this latent space, and solve a latent LQR problem to obtain timestep- and layer-specific steering signals. We provide theoretical bounds relating latent setpoint tracking to raw activation-space feature control, and empirically validate the fidelity of the reduced latent dynamics. On concept steering and video safety benchmarks, LA-LQR reduces unsafe generations relative to baselines, while preserving prompt fidelity and visual quality.

Indoor wireless propagation is governed by the interaction among three-dimensional (3D) scene geometry, radiomaterial properties, and transmitter and receiver configuration, which jointly determine both aggregate coverage behavior and path-level multipath structure. However, most learning-based site-specific prediction methods are designed for a single wireless representation, such as radiomap estimation or channel impulse response (CIR) prediction, and therefore do not explicitly exploit the propagation structure shared across heterogeneous wireless views. This paper introduces WiSER, a Wireless Scene Encoder for joint radiomap and multipath CIR prediction. WiSER maps a sparse voxel representation of an indoor scene and a transmitter location into a transmitter-conditioned sparse 3D scene memory, which is queried by two structure-aware decoders: a ray-corridor decoder for dense receiver-plane path-gain prediction and a Detection Transformer (DETR)-style set decoder for variable cardinality delay and power tap prediction. To train and evaluate this setting, we construct a co-registered indoor scene and wireless dataset pipeline using ScanNet++ indoor scenes and Sionna Ray Tracing, producing aligned sparse voxel inputs, dense radiomap labels, and unordered multipath CIR tap sets under a common coordinate frame and propagation configuration. Experimental results show that WiSER outperforms scene-specific radiomap baselines and substantially improves matched delay and power prediction over reference CIR baselines. These results suggest that transmitter-conditioned sparse 3D scene representations can serve as reusable wireless scene encoders for heterogeneous propagation queries, providing a geometry-grounded step toward representation learning and foundation-model development for AI-native wireless systems.

The Internet of Things (IoT) is a highly emerging market. It serves as a key enabler for a variety of applications like the digital twin or asset tracking in industrial scenarios. This often requires the provision of precise position information. However, systems like Global Navigation Satellite Systems (GNSS) are ruled out due to high energy costs and indoor applications. A variety of systems is discussed to close this gap. In order to contribute to the investigations of possible gold standards, this paper discusses the localization based on Low Power Wide Area Networks (LPWAN). Therefore, a concept is presented, based on Time Difference of Arrival (TDoA) measurements within the LPWAN standard ETSI TS 103 357. This paper addresses two major challenges. At first, TDoA measurements require highly precise temporal synchronization of the receiving base stations. Within this work, this issue is solved by exploiting Signals of Opportunity (SoO) as synchronization source, enabling sub-meter synchronization accuracy. A further issue arises from the Frequency Hopping (FH) waveform of the transmitting endpoints, resulting in a loss of phase information and thus usable localization bandwidth. A method is introduced to overcome this limitation. This paper states the system concept, proves its functionality in theoretical investigations and simulations. Finally, real-world measurements verify the functionality and show a 2D localization accuracy of below 10 m in Line of Sight (LOS) scenarios.

With the advent of Large Language Models, single-task and token-based multi-task models have evolved into instruction-based systems that infer task and target language implicitly from natural language prompts. This trend is reflected in IWSLT's Instruction Following Track, which this year introduced new tasks including an unknown surprise task, posing a genuine challenge against overfitting to known tasks. We present KIT's submission to the Long and Short Instruction Following tracks in the unconstrained setting. Our approach combines a general data augmentation pipeline that converts short-form corpora into long-form training data through segment concatenation, LLM-based label generation, and cross-lingual translation, yielding over 1M instances across six tasks and four languages. We further show that likelihood-based re-ranking, while highly effective for ASR, systematically degrades semantic tasks by spuriously selecting candidates generated from segmented audio processing rather than holistic long-form inference, a failure mode resolved by combining likelihood with Minimum Bayes Risk decoding.

In this paper, we present a GPU-accelerated framework for nonlinear model predictive control (NMPC) based on direct transcription and second-order interior-point methods. Many real-world systems exhibit nonlinear dynamics that cannot be accurately captured by linear models, motivating the use of NMPC. However, NMPC requires the repeated real-time solution of optimal control problems (OCP), which become computationally demanding large-scale nonlinear programs (NLPs) after transcription. Although GPU acceleration has emerged as a promising approach for nonlinear optimization, existing GPU-based NMPC workflows reconstruct structurally identical OCPs at each solve. This introduces substantial overhead even though successive solves differ only through updated system measurements or reference trajectories. To address this limitation, we introduce a parametric interior-point formulation that exploits the fixed structure of transcribed OCPs, enabling reuse of structure-dependent computations (e.g., symbolic factorization in sparse Cholesky) across re-solves. We evaluate the proposed framework on distillation column and 2D heated plate benchmarks against state-of-the-art CPU and GPU configurations. The results show that the framework achieves over an order-of-magnitude speedup in total NMPC run times. These improvements are primarily driven by reduced per-iteration solve times, with GPU execution achieving up to a 94% reduction compared to the baseline. Overall, the results demonstrate the effectiveness of exploiting repeated problem structure in GPU-accelerated NMPC and highlight the potential of the proposed framework to expand the envelope of real-time NMPC applications.

Affine frequency division multiplexing (AFDM) is a promising waveform for integrated sensing and communication (ISAC) systems owing to its superior performance in time--frequency doubly dispersive channels. However, AFDM still faces a pair of challenges: high PAPR and random data symbols produce imperfect autocorrelation sidelobes. To address these challenges, this paper proposes a real-time data-driven framework that optimizes the pre-chirp parameter $c_2$ to enhance the AFDM-ISAC performance. Specifically, a side-information-free optimization problem is formulated to reduce PAPR and the weighted integrated sidelobe levels of both aperiodic and periodic autocorrelation functions, with complexity comparable to that of the conventional AFDM receiver. Furthermore, an efficient non-monotone line-search spectral projected-gradient algorithm is developed by exploiting closed-form gradients. Simulation results demonstrate that the proposed method achieves a superior sensing vs. communications trade-off and is capable of striking a promoted bit error rate performance in the presence of severe power amplifier nonlinearity.

Automatic speech recognition systems commonly rely on reference transcriptions for evaluation, while reference-free approaches often depend on internal confidence estimation or auxiliary language models. We propose READ (Reference-free Hypothesis Evaluation with Acoustic Discrepancy), a novel metric that evaluates ASR hypotheses directly from the speech signal. READ emphasizes the acoustic grounding of hypotheses. It uses a pretrained auto-regressive TTS model to compute the conditional likelihood of speech tokens given a text hypothesis, to measure fine-grained acoustic discrepancy between speech and text. Without additional training, READ can be applied for hypothesis refinement. Experiments show that READ correlates with specific recognition errors and improves ASR outputs, achieving up to 20\% relative error rate reduction, with particularly strong gains under noisy conditions.

In two seminal articles published in 1964, Brayton and Moser introduced the concept of a mixed potential as a fundamental theoretic tool to describe and analyze a class RLC of nonlinear circuits containing resistors, capacitors and inductors. In this paper, it is shown for the first time that a mixed potential can be introduced for a class RLCM of RLC circuits containing also memristors. This is possible provided a memristor circuit is analyzed not in the traditional voltage-current domain but rather in the flux-charge domain. The flux-charge analysis method (FCAM) plays a crucial role in the extension, in particular, a key step is an equivalence principle established via FCAM between an RLCM circuit in the flux-charge domain and a nonlinear RLC circuit in the voltage-current domain. Several examples are discussed where the mixed potential is explicitly found. These include basic circuits with memristors, such as Chua's circuit with a memristor and also large-scale memristor arrays with a neural architecture. This paper is mainly devoted to the introduction of a mixed potential for memristor circuits and the study of its main theoretic properties, as the possibility to write the circuit state equations in the flux-charge domain in an effective and compact form via the mixed potential. In a companion paper [1], the mixed potential is used to obtain in a systematic way Lyapunov-like results on convergence of RLCM circuits. Those results will extend existing results on convergence that do not cover the important case where there is the simultaneous presence of capacitors and inductors in a memristor circuit.

This paper investigates joint three-dimensional (3D) trajectory planning and resource allocation for a high-altitude platform (HAPs)-unmanned aerial vehicle (UAV) bistatic integrated synthetic aperture radar (SAR) and communication (ISARAC) system in low-altitude networks. In the proposed architecture, the HAPs provides persistent wide-area connectivity by transmitting ISARAC waveforms for ground-user communications, while a low-altitude UAV exploits its proximity and mobility to passively collect ground-target echoes for high-resolution SAR imaging. We formulate a sum-rate maximization problem for ground users subject to stringent SAR imaging signal-to-noise ratio (SNR) and resolution requirements, a total energy budget for ISARAC transmission, and UAV dynamic constraints. The resulting problem is inherently nonconvex. To tackle it, an alternating optimization (AO) framework is developed, where the power-allocation subproblem with fixed UAV states admits a closed-form water-filling solution, while the UAV trajectory optimization with fixed transmit powers is handled via successive convex approximation (SCA) and difference-of-convex (DC) programming. Simulation results verify the effectiveness of the proposed approach and demonstrate its capability to jointly support persistent communication coverage and high-resolution sensing in low-altitude network scenarios.

While neural video coding (NVC) has achieved remarkable rate-distortion performance, real-time decoding on edge devices has become an important demand but remains limited by high complexity. Knowledge distillation (KD) is widely used for model acceleration, yet its application to NVC faces critical challenges. Specifically, the heterogeneity of NVC sub-modules renders uniform architectural reduction suboptimal, necessitating a per-module design for better rate-distortion-speed trade-off. However, searching for diverse architectures via existing neural architecture search (NAS) algorithms is unaffordable due to the expensive training cost of neural video codecs. Moreover, after the lightweight architecture is determined, existing distillation methods overlook the feature-energy sparsity induced by the rate-constraint, which is essential for maintaining compression performance. To address these issues, we propose a two-stage distillation framework KD-NVC. In the first stage, we introduce an acceleration-efficiency-based neural architecture search (AE-NAS) algorithm. It explores the module-wise Pareto frontier to adaptively allocate the acceleration budget across heterogeneous modules. Also, it introduces the acceleration-efficiency metric to determine the final student architecture without practically training all architecture-level candidates. In the second stage, we design an energy-aware feature distillation (EFD) loss that aligns the spatially-aggregated feature-energy signatures between the teacher and student codecs, transferring the rate-induced sparsity patterns for better compression efficiency. Experimental results demonstrate that the proposed framework consistently outperforms existing codec-oriented distillation methods, and achieves 69 FPS decoding at 1080p on RTX 5060 while maintaining comparable RD performance to VTM-LDB.

The efficiency of reactive near-field wireless power transfer (WPT) systems degrades rapidly with increasing separation distance and is highly sensitive to misalignment between transmitting and receiving coils. These limitations restrict the mobility of powered devices and confine many near-field WPT applications to static scenarios. To address these challenges, a multiple-input multiple-output (MIMO) WPT configuration is investigated due to its capability to shape the magnetic field distribution between the transmitter and receiver. Maximum power transfer efficiency can be achieved by appropriately setting the amplitude and phase of each transmitting coil; however, determining these optimal settings requires accurate knowledge of the system's S-parameters. This paper presents the use of the Nelder-Mead iterative optimization algorithm to estimate the input amplitude and phase settings that maximize transfer efficiency in a near-field WPT system. The implementation comprises a four-element transmitter and a two-element receiver. Based on measured S-parameters, the proposed approach significantly improves WPT efficiency under both aligned and misaligned conditions.