Papers

Region-of-Interest (ROI)-based image compression allocates bits unevenly according to the semantic importance of different regions. Such differentiated coding typically induces a sharp-peaked and heavy-tailed distribution. This distribution characteristic mathematically necessitates a probability model with adaptable shape parameters for accurate description. However, existing methods commonly use a Gaussian model to fit this distribution, resulting in a loss of coding performance. To systematically analyze the impact of this distribution on ROI coding, we develop a unified rate-distortion optimization theoretical paradigm. Building on this paradigm, we propose a novel Generalized Gaussian Model (GGM) to achieve flexible modeling of the latent variables distribution. To support stable optimization of GGM, we introduce effective differentiable functions and further propose a dynamic lower bound to alleviate train-test mismatch. Moreover, finite differences are introduced to solve the gradient computation after GGM fits the distribution. Experiments on COCO2017 demonstrate that our method achieves state-of-the-art in both ROI reconstruction and downstream tasks (e.g., Segmentation, Object Detection). Furthermore, compared to classical probability models, our GGM provides a more precise fit to feature distributions and achieves superior coding performance. The project page is at https://github.com/hukai-tju/ROIGGM.

Integrated sensing and communication (ISAC) is a key technology for enabling a wide range of applications in future wireless systems. However, the sensing performance is often degraded by model mismatches caused by geometric errors (e.g., position and orientation) and hardware impairments (e.g., mutual coupling and amplifier non-linearity). This paper focuses on the angle estimation performance with antenna arrays and tackles the critical challenge of array beam pattern calibration for ISAC systems. To assess calibration quality from a sensing perspective, a novel performance metric that accounts for angle estimation error, rather than beam pattern similarity, is proposed and incorporated into a differentiable loss function. Additionally, a cooperative calibration framework is introduced, allowing multiple user equipments to iteratively optimize the beam pattern based on the proposed loss functions and local data, and collaboratively update global calibration parameters. The proposed models and algorithms are validated using real-world beam pattern measurements collected in an anechoic chamber. Experimental results show that the angle estimation error can be reduced from {$\textbf{1.01}^\circ$} to $\textbf{0.11}^\circ$ in 2D calibration scenarios, and from $\textbf{5.19}^\circ$ to $\textbf{0.86}^\circ$ in 3D calibration ones.

Electric vehicle (EV) charging infrastructure introduces complex challenges to urban distribution networks, particularly under extreme demand events. A critical barrier to resilience assessment is the scale gap between micro-level charging physics and city-scale planning: minute-resolution deliverability constraints remain invisible in hourly aggregated datasets, causing purely data-driven models to exhibit non-physical behavior in high-stress regimes. This paper develops a five-stage scientific machine learning framework bridging this gap through physics-informed knowledge transfer. Stage 1 learns a temperature-pressure deliverability surface from Swiss DC fast-charging telemetry with monotonicity constraints. Stage 2 performs cross-scale injection via anchored quantile mapping. Stage 3 deploys a dual-head spatio-temporal graph neural network for joint forecasting of demand and service loss rate. Stage 4 simulates backlog dynamics under stress shocks and evaluates policy interventions. Stage 5 couples service outcomes to distribution-grid stress via transformer loading analysis. Validation on the Shenzhen UrbanEV dataset demonstrates that physics injection restores monotone stress-to-risk response (Spearman correlation coefficient equals +1.0 versus -0.8 without injection) and improves forecasting accuracy. Under a representative demand shock, the hybrid policy reduces backlog by 79.1%, restores full service within the study horizon, and limits grid stress to only 2 additional hours. The derived resilience boundary m_crit as a function of epsilon approximately equals 1.7 minus 1.0 times epsilon, providing actionable guidance linking demand flexibility to maximum absorbable stress, enabling risk-aware emergency planning under extreme events.

Audio Foundation Models (AFMs), a specialized category of Generative AI (GenAI), have the potential to transform signal processing (SP) education by integrating core applications such as speech and audio enhancement, denoising, source separation, feature extraction, automatic classification, and real-time signal analysis into learning and research. This paper introduces SPEduAFM, a conceptual AFM tailored for SP education, bridging traditional SP principles with GenAI-driven innovations. Through an envisioned case study, we outline how AFMs can enable a range of applications, including automated lecture transcription, interactive demonstrations, and inclusive learning tools, showcasing their potential to transform abstract concepts into engaging, practical experiences. This paper also addresses challenges such as ethics, explainability, and customization by highlighting dynamic, real-time auditory interactions that foster experiential and authentic learning. By presenting SPEduAFM as a forward-looking vision, we aim to inspire broader adoption of GenAI in engineering education, enhancing accessibility, engagement, and innovation in the classroom and beyond.

We address the problem of reactive motion planning for quadrotors operating in unknown environments with dynamic obstacles. Our approach leverages a 4-dimensional spatio-temporal planner, integrated with vision-based Safe Flight Corridor (SFC) generation and trajectory optimization. Unlike prior methods that rely on map fusion, our framework is mapless, enabling collision avoidance directly from perception while reducing computational overhead. Dynamic obstacles are detected and tracked using a vision-based object segmentation and tracking pipeline, allowing robust classification of static versus dynamic elements in the scene. To further enhance robustness, we introduce a backup planning module that reactively avoids dynamic obstacles when no direct path to the goal is available, mitigating the risk of collisions during deadlock situations. We validate our method extensively in both simulation and real-world hardware experiments, and benchmark it against state-of-the-art approaches, showing significant advantages for reactive UAV navigation in dynamic, unknown environments.

We propose a computationally tractable, tube-based robust nonlinear model predictive control (MPC) framework using difference-of-convex (DC) functions and sequential convex programming. For systems with differentiable discrete time dynamics, we show how to construct systematic, data-driven DC model representations using polynomials and machine learning techniques. We develop a robust tube MPC scheme that convexifies the online optimization by linearizing the concave components of the model, and we provide guarantees of recursive feasibility and robust stability. We present three data-driven procedures for computing DC models and compare performance using a planar vertical take-off and landing (PVTOL) aircraft case study.

Using multiple-input multiple-output (MIMO) with orthogonal frequency division multiplexing (OFDM) for integrated sensing and communication (ISAC) has attracted considerable attention in recent years. While most existing works focus on improving MIMO-OFDM ISAC performance, the impact of transmit power and radio-frequency (RF) circuit power consumption on energy efficiency (EE) remains relatively underexplored. To address this gap, this paper investigates joint precoding and RF chain selection for multi-user MIMO-OFDM ISAC systems, and develops energy-efficient designs for both fully digital and hybrid precoding architectures through the joint optimization of precoding and RF-chain activation. Specifically, we first formulate a novel EE maximization problem subject to sensing performance constraints. Then, efficient optimization algorithms are proposed for both architectures, together with analyses of their computational complexity and convergence behavior. Building on the proposed approaches, spectral efficiency-power consumption tradeoff designs are also provided. Simulation results demonstrate that, compared with existing schemes, the proposed approaches achieve significant improvements in the EE-sensing tradeoff for ISAC systems.

Odor-based Molecular Communication (OMC) employs odor molecules to convey information, contributing to the realization of the Internet of Everything (IoE) vision. Despite this, the practical deployment of OMC systems is currently limited by the lack of comprehensive channel models that accurately characterize particle propagation in diverse environments. While existing literature explores various aspects of molecular transport, a holistic approach that integrates theoretical modeling with experimental validation for bounded channels remains underdeveloped. In this paper, we address this gap by proposing mathematical frameworks for both bounded and unbounded OMC channels. To verify the accuracy of the proposed models, we develop a novel experimental testbed and conduct an extensive performance analysis. Our results demonstrate a strong correlation between the theoretical derivations and experimental data, providing a robust foundation for the design and analysis of future end-to-end OMC systems.

Wide-field fluorescence microscopy with compact optics often suffers from spatially varying blur due to field-dependent aberrations, vignetting, and sensor truncation, while finite sensor sampling imposes an inherent trade-off between field of view (FOV) and resolution. Computational Miniaturized Mesoscope (CM2) alleviate the sampling limit by multiplexing multiple sub-views onto a single sensor, but introduce view crosstalk and a highly ill-conditioned inverse problem compounded by spatially variant point spread functions (PSFs). Prior learning-based spatially varying (SV) reconstruction methods typically rely on global SV operators with fixed input sizes, resulting in memory and training costs that scale poorly with image dimensions. We propose SV-CoDe (Spatially Varying Coordinate-conditioned Deconvolution), a scalable deep learning framework that achieves uniform, high-resolution reconstruction across a 6.5 mm FOV. Unlike conventional methods, SV-CoDe employs coordinate-conditioned convolutions to locally adapt reconstruction kernels; this enables patch-based training that decouples parameter count from FOV size. SV-CoDe achieves the best image quality in both simulated and experimental measurements while requiring 10x less model size and 10x less training data than prior baselines. Trained purely on physics-based simulations, the network robustly generalizes to bead phantoms, weakly scattering brain slices, and freely moving C. elegans. SV-CoDe offers a scalable, physics-aware solution for correcting SV blur in compact optical systems and is readily extendable to a broad range of biomedical imaging applications.

Existing generative models for unsupervised anomalous sound detection are limited by their inability to fully capture the complex feature distribution of normal sounds, while the potential of powerful diffusion models in this domain remains largely unexplored. To address this challenge, we propose a novel framework, TLDiffGAN, which consists of two complementary branches. One branch incorporates a latent diffusion model into the GAN generator for adversarial training, thereby making the discriminator's task more challenging and improving the quality of generated samples. The other branch leverages pretrained audio model encoders to extract features directly from raw audio waveforms for auxiliary discrimination. This framework effectively captures feature representations of normal sounds from both raw audio and Mel spectrograms. Moreover, we introduce a TMixup spectrogram augmentation technique to enhance sensitivity to subtle and localized temporal patterns that are often overlooked. Extensive experiments on the DCASE 2020 Challenge Task 2 dataset demonstrate the superior detection performance of TLDiffGAN, as well as its strong capability in anomalous time-frequency localization.

Audio deepfakes generated by modern TTS and voice conversion systems are increasingly difficult to distinguish from real speech, raising serious risks for security and online trust. While state-of-the-art self-supervised models provide rich multi-layer representations, existing detectors treat layers independently and overlook temporal and hierarchical dependencies critical for identifying synthetic artefacts. We propose HierCon, a hierarchical layer attention framework combined with margin-based contrastive learning that models dependencies across temporal frames, neighbouring layers, and layer groups, while encouraging domain-invariant embeddings. Evaluated on ASVspoof 2021 DF and In-the-Wild datasets, our method achieves state-of-the-art performance (1.93% and 6.87% EER), improving over independent layer weighting by 36.6% and 22.5% respectively. The results and attention visualisations confirm that hierarchical modelling enhances generalisation to cross-domain generation techniques and recording conditions.

This work presents the first systematic investigation of speech bias in multilingual MLLMs. We construct and release the BiasInEar dataset, a speech-augmented benchmark based on Global MMLU Lite, spanning English, Chinese, and Korean, balanced by gender and accent, and totaling 70.8 hours ($\approx$4,249 minutes) of speech with 11,200 questions. Using four complementary metrics (accuracy, entropy, APES, and Fleiss' $\kappa$), we evaluate nine representative models under linguistic (language and accent), demographic (gender), and structural (option order) perturbations. Our findings reveal that MLLMs are relatively robust to demographic factors but highly sensitive to language and option order, suggesting that speech can amplify existing structural biases. Moreover, architectural design and reasoning strategy substantially affect robustness across languages. Overall, this study establishes a unified framework for assessing fairness and robustness in speech-integrated LLMs, bridging the gap between text- and speech-based evaluation. The resources can be found at https://github.com/ntunlplab/BiasInEar.

The rapid growth of hyperscale data centers driven by Large Language Models and Artificial Intelligence workloads has introduced new challenges for power systems. These facilities experience abrupt power variations during model training and check-point-saving events, causing voltage deviations and frequency disturbances. Moreover, they operate as passive loads that draw power without offering any grid support. This paper presents an integrated architecture that combines Battery Energy Storage Systems (BESSs) within data centers using Grid-Forming inverters to provide active grid-support functions. Simulation results through MATLAB/Simulink demonstrate accurate power reference tracking under dynamic loading, with eight coordinated BESS units supplying instantaneous power during training and saving conditions. Under single-phase voltage depression near the data center bus, the BESS delivered reactive power support similar to a Static Synchronous Compensator. During grid disconnection, seamless islanded operation was achieved with stable voltage, frequency, and continuous power delivery at the data center bus.

Recent speech foundation models excel at multilingual automatic speech recognition (ASR) for high-resource languages, but adapting them to low-resource languages remains challenging due to data scarcity and efficiency constraints. Full-model fine-tuning is computationally expensive and prone to overfitting, while parameter-efficient methods like LoRA apply adaptation uniformly across layers, overlooking internal representations thus compromising effectiveness and efficiency. We analyze multilingual ASR models and reveal a U-shaped adaptability pattern: early and late layers are language-specific and require more adaptation, while intermediate layers retain shared semantics and need less. Building on this observation, we propose DAMA, a Depth-Aware Model Adaptation framework that allocates adaptation capacity according to each layer's role. DAMA also introduces Singular Value Decomposition (SVD)-based initialization to constrain adaptation and preserve the U-shaped pattern, as well as a frozen middle-layer basis for further efficiency. Evaluated on 18 low-resource languages across two benchmark datasets, DAMA matches or surpasses state-of-the-art accuracy with 80% fewer trainable parameters, achieves a 29% error reduction under extreme data scarcity, and significantly improves memory, training time, and computational efficiency over baselines. These results highlight the benefits of structure-aware adaptation for efficient, scalable multilingual ASR.

Objective: Thyroid ultrasound is commonly performed using a combination of static images and cine clips (video recordings). However, the exact utility and impact of cine images remains unknown. This study aimed to evaluate the impact of cine imaging on accuracy and consistency of thyroid nodule assessment, using the American College of Radiology Thyroid Imaging, Reporting and Data System (ACR TI-RADS). Methods: 50 benign and 50 malignant thyroid nodules with cytopathology results, from 89 patients, were included. A reader study with 4 specialty-trained radiologists was then conducted over 3 rounds, assessing only static images in the first two rounds and both static and cine images in the third round. TI-RADS scores and the consequent management recommendations were then evaluated by comparing them to the malignancy status of the nodules. Results: Mean sensitivity for malignancy detection was 0.65 for static images and 0.67 with both static and cine images (p=0.41). Specificity was 0.20 for static images and 0.22 with both static and cine images (p=0.25). Management recommendations were similar with and without cine images. Intrareader agreement on feature assignments was slightly lower when cine images were added (0.83 vs 0.87, p<0.001), and TI-RADS point totals changed more between rounds with cine images. Conclusion: The inclusion of cine imaging for thyroid nodule assessment on ultrasound did not significantly change diagnostic performance in this reader study, although it changed some feature assignments. Current practice guidelines, which do not mandate cine imaging, may be sufficient for TI-RADS-based management recommendations.

We address the tracking problem for a class of uncertain non-affine nonlinear systems with high relative degrees, performing non-repetitive tasks. We propose a rigorously proven, robust adaptive learning control scheme that relies on a gradient descent parameter adaptation law to handle the unknown time-varying parameters of the system, along with a state estimator that estimates the unmeasurable state variables. Furthermore, despite the inherently complex nature of the non-affine system, we provide an explicit iterative computation method to facilitate the implementation of the proposed control scheme. The paper includes a thorough analysis of the performance of the proposed control strategy, and simulation results are presented to demonstrate the effectiveness of the approach.

Power system operators routinely perform N-1 contingency analysis, yet conventional tools provide limited guidance on which lines or transformers deserve heightened attention during fast post-fault transients. In particular, static screening does not reveal whether (1) the same faulted line repeatedly triggers severe downstream overloads, or (2) a specific transformer emerges as vulnerable across many distinct fault scenarios. This paper introduces a real-time dynamic N-1 screening framework that addresses this gap by estimating, for each counterfactual single-phase transmission fault, the probability of transient overcurrent on critical grid elements. The output is an operator-facing dashboard that ranks (a) faulted lines whose outages most frequently lead to dangerous transformer overloads, and (b) transformers that consistently overload across top-risk scenarios, both of which are actionable indicators for real-time situational awareness. The approach models post-fault electromechanical dynamics using a linear stochastic formulation of the swing equations with short-lived, fault-localized uncertainty, and combines analytic transient evaluation with cross-entropy based importance sampling to efficiently estimate rare but high-impact events. All N-1 contingencies are evaluated in parallel with linear computational complexity. The framework is demonstrated on the IEEE 118-bus system, where it reveals latent high-risk lines and transformers that remain invisible under deterministic dynamic or static N-1 analysis. Results show orders-of-magnitude computational speedup relative to brute-force Monte Carlo, enabling practical deployment within real-time operational cycles.

Understanding how environmental and operational conditions influence vessel speed is crucial for characterizing navigational conditions in the Arctic. We analyzed Automatic Identification System (AIS) data from 2010-2019 to examine vessel speed over ground (SOG). Over half of the AIS records showed zero SOG, and treating zero and positive SOG as a single continuous process can obscure important patterns. We therefore applied a two-stage machine learning framework, first modeling the probability of SOG greater than zero and then modeling SOG conditional on being positive. AIS observations were integrated with sea ice concentration, course over ground, wind, bathymetric depth, distance to coast, vessel group, and navigational status. Gradient boosted decision trees with random effects captured nonlinear environmental responses while accounting for repeated observations. The positive SOG classifier achieved strong discrimination (AUC = 0.85), while the conditional speed model explained approximately 77 percent of out-of-fold variance. SHAP values quantified covariate effects by decomposing model predictions into additive contributions from individual variables. Distance to coast and bathymetric depth were dominant determinants of both the likelihood and magnitude of vessel speed, while changes in course, vessel group, and navigational status introduced secondary variation. Wind and sea ice effects were modest. Together, these results empirically characterize Arctic vessel operating regimes relevant to speed management and corridor-level assessment.

Accurate currency recognition is essential for assistive technologies, particularly for visually impaired individuals who rely on others to identify banknotes. This dependency puts them at risk of fraud and exploitation. To address these challenges, we first build a new Bangladeshi banknote dataset that includes both controlled and real-world scenarios, ensuring a more comprehensive and diverse representation. Next, to enhance the dataset's robustness, we incorporate four additional datasets, including public benchmarks, to cover various complexities and improve the model's generalization. To overcome the limitations of current recognition models, we propose a novel hybrid CNN architecture that combines MobileNetV3-Large and EfficientNetB0 for efficient feature extraction. This is followed by an effective multilayer perceptron (MLP) classifier to improve performance while keeping computational costs low, making the system suitable for resource-constrained devices. The experimental results show that the proposed model achieves 97.95% accuracy on controlled datasets, 92.84% on complex backgrounds, and 94.98% accuracy when combining all datasets. The model's performance is thoroughly evaluated using five-fold cross-validation and seven metrics: accuracy, precision, recall, F1-score, Cohen's Kappa, MCC, and AUC. Additionally, explainable AI methods like LIME and SHAP are incorporated to enhance transparency and interpretability.

As industrial manufacturing scales, automating fine-grained product image analysis has become critical for quality control. However, existing approaches are hindered by limited dataset coverage and poor model generalization across diverse and complex anomaly patterns. To address these challenges, we introduce MAU-Set, a comprehensive dataset for Multi-type industrial Anomaly Understanding. It spans multiple industrial domains and features a hierarchical task structure, ranging from binary classification to complex reasoning. Alongside this dataset, we establish a rigorous evaluation protocol to facilitate fair and comprehensive model assessment. Building upon this foundation, we further present MAU-GPT, a domain-adapted multimodal large model specifically designed for industrial anomaly understanding. It incorporates a novel AMoE-LoRA mechanism that unifies anomaly-aware and generalist experts adaptation, enhancing both understanding and reasoning across diverse defect classes. Extensive experiments show that MAU-GPT consistently outperforms prior state-of-the-art methods across all domains, demonstrating strong potential for scalable and automated industrial inspection.

Unsupervised anomaly detection (AD) in medical images aims to identify abnormal regions without relying on pixel-level annotations, which is crucial for scalable and label-efficient diagnostic systems. In this paper, we propose a novel anomaly detection framework based on DINO-V3 representations, termed DINO-AD, which leverages self-supervised visual features for precise and interpretable anomaly localization. Specifically, we introduce an embedding similarity matching strategy to select a semantically aligned support image and a foreground-aware K-means clustering module to model the distribution of normal features. Anomaly maps are then computed by comparing the query features with clustered normal embeddings through cosine similarity. Experimental results on both the Brain and Liver datasets demonstrate that our method achieves superior quantitative performance compared with state-of-the-art approaches, achieving AUROC scores of up to 98.71. Qualitative results further confirm that our framework produces clearer and more accurate anomaly localization. Extensive ablation studies validate the effectiveness of each proposed component, highlighting the robustness and generalizability of our approach.

The digitization of agricultural advisory services in India requires robust Automatic Speech Recognition (ASR) systems capable of accurately transcribing domain-specific terminology in multiple Indian languages. This paper presents a benchmarking framework for evaluating ASR performance in agricultural contexts across Hindi, Telugu, and Odia languages. We introduce evaluation metrics including Agriculture Weighted Word Error Rate (AWWER) and domain-specific utility scoring to complement traditional metrics. Our evaluation of 10,934 audio recordings, each transcribed by up to 10 ASR models, reveals performance variations across languages and models, with Hindi achieving the best overall performance (WER: 16.2%) while Odia presents the greatest challenges (best WER: 35.1%, achieved only with speaker diarization). We characterize audio quality challenges inherent to real-world agricultural field recordings and demonstrate that speaker diarization with best-speaker selection can substantially reduce WER for multi-speaker recordings (upto 66% depending on the proportion of multi-speaker audio). We identify recurring error patterns in agricultural terminology and provide practical recommendations for improving ASR systems in low-resource agricultural domains. The study establishes baseline benchmarks for future agricultural ASR development.

Arctic sea ice plays a critical role in regulating Earth's climate system, significantly influencing polar ecological stability and human activities in coastal regions. Recent advances in artificial intelligence have facilitated the development of skillful pan-Arctic sea ice forecasting systems, where data-driven approaches showcase tremendous potential to outperform conventional physics-based numerical models in terms of accuracy, computational efficiency and forecasting lead times. Despite the latest progress made by deep learning (DL) forecasting models, most of their skillful forecasting lead times are confined to daily subseasonal scale and monthly averaged values for up to six months, which drastically hinders their deployment for real-world applications, e.g., maritime routine planning for Arctic transportation and scientific investigation. Extending daily forecasts from subseasonal to seasonal (S2S) scale is scientifically crucial for operational applications. To bridge the gap between the forecasting lead time of current DL models and the significant daily S2S scale, we introduce IceBench-S2S, the first comprehensive benchmark for evaluating DL approaches in mitigating the challenge of forecasting Arctic sea ice concentration in successive 180-day periods. It proposes a generalized framework that first compresses spatial features of daily sea ice data into a deep latent space. The temporally concatenated deep features are subsequently modeled by DL-based forecasting backbones to predict the sea ice variation at S2S scale. IceBench-S2S provides a unified training and evaluation pipeline for different backbones, along with practical guidance for model selection in polar environmental monitoring tasks.

Precision in radiation therapy relies on immobilization systems that limit patient motion. Thermoplastic masks are commonly used for this purpose, but subtle voluntary and involuntary movements such as jaw shifts, deep breathing, or eye squinting may still compromise treatment accuracy. Existing motion tracking methods are limited: optical systems require a clear line of sight and only detect surface motion, while X-ray-based tracking introduces additional ionizing radiation. This study explores the use of low-power, non-ionizing millimeter-wave (mmWave) sensing for through-mask motion detection. We characterize the RF properties of thermoplastic mask material in the 28-38 GHz range and perform motion detection using a 1 GHz bandwidth centered at 28 GHz. We use a frequency-domain system with horn antennas in a custom-built anechoic chamber to capture changes in the amplitude and phase of transmitted RF waves in response to subtle head and facial movements. These findings lay groundwork for future real-time through-mask motion tracking and future integration with multi-antenna systems and machine learning for error correction during radiotherapy.

We present a lightweight multimodal baseline for emotion recognition in conversations using the SemEval-2024 Task 3 dataset built from the sitcom Friends. The goal of this report is not to propose a novel state-of-the-art method, but to document an accessible reference implementation that combines (i) a transformer-based text classifier and (ii) a self-supervised speech representation model, with a simple late-fusion ensemble. We report the baseline setup and empirical results obtained under a limited training protocol, highlighting when multimodal fusion improves over unimodal models. This preprint is provided for transparency and to support future, more rigorous comparisons.

Total energy shaping through interconnection and damping assignment passivity-based control (IDA-PBC) provides a powerful and systematic framework for stabilizing underactuated mechanical systems. Despite its theoretical appeal, incorporating actuator limitations into total energy shaping remains a largely open problem, with only limited results reported in the existing literature. In practice, the closed-loop behavior of energy-shaping controllers is strongly affected by the kinetic energy shaping terms. In this paper, a simultaneous IDA-PBC (SIDA-PBC) framework is employed to systematically attenuate the kinetic energy shaping terms by exploiting generalized forces, without altering the matching partial differential equations (PDEs). The free component of the generalized forces is derived analytically via an $\ell_\infty$-norm optimization formulation. Although a reduction in kinetic energy shaping terms does not necessarily guarantee a decrease in the overall control effort, the proposed approach effectively suppresses kinetic energy shaping components and achieves a reduced control magnitude whenever such a reduction is structurally feasible. Unlike existing approaches based on gyroscopic terms, which require multiple actuators, the proposed method is applicable to mechanical systems with a single actuator. Simulation and experimental results are provided to validate the effectiveness of the proposed approach.

Although the stabilization of underactuated systems remains a challenging problem, the total energy shaping approach provides a general framework for addressing this objective. However, the practical implementation of this method is hindered by the need to analytically solve a set of partial differential equations (PDEs), which constitutes a major obstacle. In this paper, a rotary inverted pendulum system is considered, and an interconnection and damping assignment passivity-based control (IDA-PBC) scheme is developed by deriving concise analytical solutions to the kinetic and potential energy PDEs. Furthermore, a novel robust term is incorporated into the control law to compensate for a specific class of disturbances that has not been addressed within the existing IDA-PBC literature. The effectiveness of the proposed method is validated through numerical simulations, demonstrating satisfactory control performance.

While point cloud-based applications are gaining traction due to their ability to provide rich and immersive experiences, they critically need efficient coding solutions due to the large volume of data involved, often many millions of points per object. The JPEG Pleno Learning-based Point Cloud Coding standard, as the first learning-based coding standard for static point clouds, has set a foundational framework with very competitive compression performance regarding the relevant conventional and learning-based alternative point cloud coding solutions. This paper proposes a novel lightweight point cloud geometry coding model that significantly reduces the complexity of the standard, which is essential for the broad adoption of this coding standard, particularly in resource-constrained environments, while simultaneously achieving small average compression efficiency benefits. The novel coding model is based on the pioneering adoption of a compressed domain approach for the super-resolution model, in addition to a major reduction of the number of latent channels. A reduction of approximately 70% in the total number of model parameters is achieved while simultaneously offering slight average compression performance gains for the JPEG Pleno Point Cloud coding dataset.

Autonomous Underwater Vehicles (AUVs) are a highly promising technology for ocean exploration and diverse offshore operations, yet their practical deployment is constrained by energy efficiency and endurance. To address this, we propose Current-Harnessing Stage-Gated MPC, which exploits ocean currents via a per-stage scalar which indicates the "helpfulness" of ocean currents. This scalar is computed along the prediction horizon to gate lightweight cost terms only where the ocean currents truly aids the control goal. The proposed cost terms, that are merged in the objective function, are (i) a Monotone Cost Shaping (MCS) term, a help-gated, non-worsening modification that relaxes along-track position error and provides a bounded translational energy rebate, guaranteeing the shaped objective is never larger than a set baseline, and (ii) a speed-to-fly (STF) cost component that increases the price of thrust and softly matches ground velocity to the ocean current, enabling near zero water-relative "gliding". All terms are C1 and integrate as a plug-and-play in MPC designs. Extensive simulations with the BlueROV2 model under realistic ocean current fields show that the proposed approach achieves substantially lower energy consumption than conventional predictive control while maintaining comparable arrival times and constraint satisfaction.

The Internet has evolved through successive architectural abstractions that enabled unprecedented scale, interoperability, and innovation. Packet-based networking enabled the reliable transport of bits; cloud-native systems enabled the orchestration of distributed computation. Today, the emergence of autonomous, learning-based systems introduces a new architectural challenge: intelligence is increasingly embedded directly into network control, computation, and decision-making, yet the Internet lacks a structural foundation for representing and exchanging meaning. In this paper, we argue that cognition alone: pattern recognition, prediction, and optimization, is insufficient for the next generation of networked systems. As autonomous agents act across safety-critical and socio-technical domains, systems must not only compute and communicate, but also comprehend intent, context, and consequence. We introduce the concept of a Semantic Layer: a new architectural stratum that treats meaning as a first-class construct, enabling interpretive alignment, semantic accountability, and intelligible autonomous behavior. We show that this evolution leads naturally to a Syntactic-Semantic Internet. The syntactic stack continues to transport bits, packets, and workloads with speed and reliability, while a parallel semantic stack transports meaning, grounding, and consequence. We describe the structure of this semantic stack-semantic communication, a semantic substrate, and an emerging Agentic Web, and draw explicit architectural parallels to TCP/IP and the World Wide Web. Finally, we examine current industry efforts, identify critical architectural gaps, and outline the engineering challenges required to make semantic interoperability a global, interoperable infrastructure.