321 papers
This paper establishes conditions under which multi-stage Runge-Kutta methods preserve strong contractivity for infinitesimally contracting continuous-time systems, deriving coefficient-dependent criteria for explicit schemes and extending classical implicit guarantees to strong contractivity across , , and norms while ensuring unique solvability via an auxiliary dynamic system.
The paper proposes Fair-Gate, a fairness-aware and interpretable risk-gating framework that mitigates sex-related performance gaps in voice biometrics by addressing demographic shortcut learning and feature entanglement through risk extrapolation and a local complementary routing mechanism, thereby improving the utility-fairness trade-off on the VoxCeleb1 dataset.
This paper proposes RHOSI, an efficient resource allocation framework that leverages a Reconfigurable Holographic Surface (RHS)-aided UAV to jointly optimize beamforming, phase shifts, and spatial deployment, thereby significantly enhancing the anti-jamming resilience and throughput of Integrated Sensing and Communication (ISAC) systems.
This paper demonstrates that continued pretraining of wav2vec2-bert-2.0 using unlabeled audio and limited labeled data achieves state-of-the-art Swahili automatic speech recognition with a 3.24% word error rate, representing a 61% relative improvement over the previous best academic system.
This paper proposes SliceFed, a novel Federated Constrained Multi-Agent Deep Reinforcement Learning framework that leverages a Lagrangian primal-dual approach with Proximal Policy Optimization to optimize dynamic spectrum slicing in 6G networks, achieving near-perfect URLLC latency compliance and robust interference management while preserving data privacy through federated learning.
This paper introduces AnimeScore, a preference-based framework and dataset that establishes a standardized objective metric for evaluating "anime-like" speech by leveraging pairwise rankings and SSL-based models to overcome the limitations of traditional subjective assessments.
This paper proposes a low-complexity iterative algorithm for an ISAC-enabled multi-UAV collaborative target sensing scheme that jointly optimizes UAV-BS association, trajectories, and bandwidth allocation to minimize the posterior Cramer-Rao bound for tracking dynamic low-altitude targets while satisfying communication constraints.
This paper proposes a robust Standard Condition Number (SCN)-based detection framework for MIMO ISAC systems under noise uncertainty, deriving closed-form performance expressions via random matrix theory and optimizing power allocation to outperform conventional detectors in interference-prone environments.
This paper proposes a hybrid zonotope-based framework for computing exact and scalable forward and backward reachable sets of closed-loop ReLU RNNs without unrolling, featuring a tunable relaxation scheme to balance computational complexity and approximation accuracy while providing a sufficient condition for safety certification.
This paper proposes a novel free-space optical communication system using skyrmion number modulation, which encodes data onto topological invariants and employs an intensity-based masking technique to achieve robust performance against atmospheric turbulence.
This paper proposes a multi-agent deep reinforcement learning framework utilizing virtual anchor nodes for unmanned aerial vehicles to achieve superior accuracy and efficiency in localizing chemical plume sources from undocumented orphaned wells compared to traditional fluxotaxis methods.
This paper proposes a learnable template-matching approach integrated with an electromagnetic signal feature embedding and clutter filtering mechanism to effectively suppress environmental clutter and enhance the accuracy of micro-deformation monitoring in Integrated Sensing and Communication (ISAC) platforms.
This paper presents a real-world case study demonstrating how machine learning models can analyze core-making process data in heavy vehicle manufacturing to proactively predict and prevent casting defects, thereby shifting quality control from reactive to proactive and improving overall production efficiency.
SEMamba++ is a computationally efficient speech restoration framework that outperforms existing baselines by introducing a Frequency GLP block and a multi-resolution parallel time-frequency dual-processing mechanism to better capture global, local, and periodic spectral patterns inherent in speech.
The paper proposes Relativistic Adversarial Feedback (RAF), a novel training objective that integrates speech self-supervised learning and relativistic pairing to significantly enhance the fidelity and generalization of GAN-based vocoders, enabling a lightweight BigVGAN variant to outperform larger models with superior perceptual quality.
This paper introduces a new dataset and demonstrates that surface electromyography (sEMG) signals from facial and neck muscles can reliably decode affective states, particularly frustration, during both phonated and silent speech, highlighting their potential for affect-aware silent speech interfaces.
This paper proposes a novel rotatable antenna (RA)-enabled covert communication system that jointly optimizes transmit beamforming and RA rotational angles via an alternating optimization algorithm to maximize covert rate while satisfying covertness and power constraints, demonstrating superior performance over benchmark schemes.
This paper presents an optimized C++ implementation of the QPALM-OCP algorithm that exploits the inherent stage-wise parallelism in optimal control problems to significantly enhance computational performance through parallelization and vectorization.
This paper derives asymptotically exact closed-form formulas for the high signal-to-noise ratio probability in continuous fluid antenna systems operating over Rayleigh fading channels with spatially coherent isotropic correlation, ultimately establishing scaling laws that describe performance gains across one, two, and three-dimensional spaces and identifying optimal system dimensions.
This paper proposes and validates a joint optimization framework for uplink PD-NOMA systems assisted by continuous reconfigurable intelligent surfaces (CRIS) that derives an accurate bit error rate (BER) expression under spatially correlated fading to eliminate BER floors and outperform conventional OMA and non-optimized NOMA schemes.