arXiv subjects

Yuhan Su

Publications and source records attributed to Yuhan Su.

12 recordsLinked to original sources

arXiv2026

LSEAD: A Privacy-Preserving LLM-Based Speech Analysis Framework for Early Alzheimer's Disease Screening

Early diagnosis of Alzheimer's disease (AD) is critical for enabling timely interventions that may slow disease progression and improve patient outcomes. There is a growing need for AD detection methods that are non-invasive and cost-effective, especially in real-world clinical settings with diverse patient populations and recording conditions. Speech-based screening addresses these needs by using natural speech collected without specialized equipment. Recent advances in large language models (LLMs) have improved speech analysis by providing rich linguistic representations and strong generalization. In this study, we propose LSEAD, a speech-based AD detection framework using pretrained open-source LLMs. Speech recordings are automatically transcribed, and text embeddings are extracted using locally deployed LLMs. Principal component analysis (PCA) is applied to reduce dimensionality before classification. Because the framework relies only on speech transcripts and locally deployed models, it supports privacy-preserving AD risk assessment without external data exchange. We evaluate LSEAD on the ADReSS20 and ADReSSo2021 benchmark datasets. Experimental results show that LLM-based embeddings generalize well across datasets and improve AD classification accuracy by up to 5 percent over existing methods, especially for early-stage detection. These results demonstrate that LSEAD provides a practical, secure, and scalable approach for early AD screening.

eess.AS↗

arXiv2026

Flying over The Uncertain Nature (FORTUNE): Intelligent and Humanistic 3D Path Planning for Low-Altitude Collaboration

The proliferation of low-altitude intelligent agents is increasing the demand for timely and socially responsible collaborative sensing in dynamic urban environments. However, jointly addressing heterogeneous spatiotemporal demands, environmental uncertainty, and human-centered operational constraints remains challenging. This paper studies 3D multi-UAV path planning and task assignment under uncertain ground PoI demands. Unlike existing work assuming static and fully known PoIs, we model persistent, temporally predictable, and emergent demands within a unified framework. We further incorporate altitude-dependent societal and environmental costs, including noise exposure and public safety risks, to balance sensing performance with socially compliant operations. To solve the resulting large-scale mixed-integer nonlinear problem, we propose FORTUNE, a hierarchical offline-online framework. Offline, a Transformer predicts Type-II PoI activation windows, while an enhanced sparrow search algorithm generates coordinated flight plans through priority-aware decoding and danger-aware evolution. Online, a lightweight refinement module accommodates emerging Type-III PoIs while preserving global mission coherence. Experiments on real-world traffic data and synthetic scenarios show that FORTUNE consistently outperforms state-of-the-art methods in effectiveness, scalability, and practical applicability.

cs.RO↗

arXiv2026

CultureVidBench: Benchmarking Cultural Understanding in Text-to-Video Generation

Text-to-video (T2V) generation models have advanced rapidly, yet their ability to represent diverse cultural contexts remains underexplored. Existing benchmarks mainly focus on perceptual quality, physical plausibility, and text-video alignment, but do not directly assess whether generated videos capture culturally specific objects, actions, rituals, visible text, or audio cues. We introduce CultureVidBench, a comprehensive benchmark for evaluating cultural understanding in T2V generation. CultureVidBench contains 1,000 curated prompts covering 12 countries, 6 continents, 8 cultural regions, and 14 cultural aspects organized into three categories: material culture, social practice & performance, and ritual & ceremony. Designed specifically for video generation, CultureVidBench emphasizes dynamic and multimodal cultural representation, including social interactions, ritual procedure, and culturally appropriate visible text and audio. We evaluate seven representative T2V models through human user studies and MLLM-based automatic assessment across cultural faithfulness, multimodal cultural rendering, semantic adherence, and perceptual quality. Results show that although current models achieve strong semantic adherence and visual quality, they often fail to faithfully capture fine-grained cultural details, particularly for underrepresented regions, rituals, and multimodal cultural cues.

cs.CV↗

arXiv2026

Toward Generalizable Cognitive Impairment Detection with Speech-Based Multimodal Large Language Models

Cognitive impairment (CI) is a growing public health concern. Early and accurate diagnosis is critical for enabling timely intervention and improving patient outcomes. Speech-based CI detection has emerged as a promising non-invasive approach, as speech signals encode both linguistic and acoustic markers associated with cognitive decline. Recent advances in large language models (LLMs) further strengthen the potential of speech-based assessment by enabling more expressive representation learning and improved generalization across diverse speakers, recording devices, and clinical environments. Moreover, multimodal learning by jointly modeling linguistic and acoustic features allows for a more comprehensive characterization of cognitive and behavioral changes related to CI, leading to more reliable detection. In this work, we propose a multimodal CI detection framework based on open-source LLMs that integrates speech audio and corresponding transcripts while preserving patient privacy. Acoustic embeddings are extracted directly from speech signals, while textual embeddings are generated from automatically transcribed speech. These modality-specific embeddings are then concatenated to create a combined feature vector and used for downstream classification, without requiring access to raw or sensitive patient data. The proposed approach is evaluated on the ADReSS20 and ADReSSo21 benchmark datasets. Experimental results show that the proposed multimodal framework achieves an CI classification accuracy of 92.4% and consistently outperforms single-modality baselines. Our work establishes a new state-of-the-art for CI identification, with the proposed method demonstrating superior cross-dataset generalization. This advance highlights the power of an LLM-based multimodal framework that fuses linguistic and acoustic data to enable robust, scalable, and non-invasive screening.

eess.SP↗

arXiv2026

HermesHFL: Incentive-Compatible Hierarchical Federated Unlearning for Dynamic LLM Fine-Tuning

Hierarchical federated unlearning (HFUL) for large language model (LLM) fine-tuning faces significant challenges due to hierarchical aggregation, dynamic client participation, and strong parameter coupling in LLM adaptation. Selectively removing client contributions is particularly difficult because model updates propagate across multiple aggregation stages while unlearning requests may coincide with client departures and rejoining. To address these issues, we propose HermesHFL, a hierarchical federated learning framework that supports selective unlearning, dynamic client participation, and client reintegration for scalable LLM fine-tuning via parameter-efficient fine-tuning (PEFT) with LoRA. We formulate a unified optimization problem that jointly models client participation, edge association, incentive allocation, and unlearning under heterogeneous client behaviors. To solve this problem efficiently, we develop Neogen, a neural-guided bilevel evolutionary optimization framework that combines CMA-ES for continuous incentive optimization with a CHC-based evolutionary mechanism for discrete participation and association decisions. A neural surrogate further accelerates optimization and improves search efficiency. Extensive experiments on LLM fine-tuning tasks demonstrate that HermesHFL consistently outperforms state-of-the-art baselines in model utility, unlearning effectiveness, convergence stability, and resource efficiency.

cs.CE↗

arXiv2026

Forecasting-Driven Stable Successor Matching for UAV-Assisted Continuous Edge Services

Continuous and reliable service support is crucial for emerging latency-sensitive and computation-intensive applications in UAV-assisted edge networks (UENs) due to operational dynamics and environmental uncertainty. Although conventional designs can improve coverage and computing efficiency, they often rely on instantaneous resource optimization or reactive handover, rendering ongoing services vulnerable to non-negligible interruptions when the serving UAV degrades due to mobility, energy depletion, or channel dynamics. To avoid such post-failure recovery, a promising approach is to prepare a successor UAV in advance, i.e., a standby UAV that reserves minimal resources and synchronizes service context for possible takeover. Thus, we consider a dynamic UEN architecture where each mobile user carries an ongoing computing mission requiring persistent service support, while UAVs provide wireless access and computing services under time-varying network dynamics and stringent onboard energy constraints. To facilitate proactive and continuous service provisioning, we propose a forecasting-driven proactive reservation-based continuous service scheduling framework, termed Fresco. In Fresco, an LSTM-based module is first used to predict short-term disruption risks of ongoing missions from historical network observations. Guided by these predictions, an online risk-aware successor matching scheme selects suitable standby UAVs for high-risk missions under delay, resource, and energy constraints, while incorporating minimal communication/computation reservation and lightweight service-context synchronization for efficient takeover preparation. Experiments show that Fresco significantly reduces service interruptions and improves mission continuity over reactive and non-predictive baselines, with only modest reservation overhead.

cs.NI↗

arXiv2025

Adaptive UAV-Assisted Hierarchical Federated Learning: Optimizing Energy, Latency, and Resilience for Dynamic Smart IoT

Hierarchical Federated Learning (HFL) extends conventional Federated Learning (FL) by introducing intermediate aggregation layers, enabling distributed learning in geographically dispersed environments, particularly relevant for smart IoT systems, such as remote monitoring and battlefield operations, where cellular connectivity is limited. In these scenarios, UAVs serve as mobile aggregators, dynamically connecting terrestrial IoT devices. This paper investigates an HFL architecture with energy-constrained, dynamically deployed UAVs prone to communication disruptions. We propose a novel approach to minimize global training costs by formulating a joint optimization problem that integrates learning configuration, bandwidth allocation, and device-to-UAV association, ensuring timely global aggregation before UAV disconnections and redeployments. The problem accounts for dynamic IoT devices and intermittent UAV connectivity and is NP-hard. To tackle this, we decompose it into three subproblems: \textit{(i)} optimizing learning configuration and bandwidth allocation via an augmented Lagrangian to reduce training costs; \textit{(ii)} introducing a device fitness score based on data heterogeneity (via Kullback-Leibler divergence), device-to-UAV proximity, and computational resources, using a TD3-based algorithm for adaptive device-to-UAV assignment; \textit{(iii)} developing a low-complexity two-stage greedy strategy for UAV redeployment and global aggregator selection, ensuring efficient aggregation despite UAV disconnections. Experiments on diverse real-world datasets validate the approach, demonstrating cost reduction and robust performance under communication disruptions.

cs.LG↗

arXiv2025

Towards Seamless Hierarchical Federated Learning under Intermittent Client Participation: A Stagewise Decision-Making Methodology

Federated Learning (FL) offers a pioneering distributed learning paradigm that enables devices/clients to build a shared global model. This global model is obtained through frequent model transmissions between clients and a central server, which may cause high latency, energy consumption, and congestion over backhaul links. To overcome these drawbacks, Hierarchical Federated Learning (HFL) has emerged, which organizes clients into multiple clusters and utilizes edge nodes (e.g., edge servers) for intermediate model aggregations between clients and the central server. Current research on HFL mainly focus on enhancing model accuracy, latency, and energy consumption in scenarios with a stable/fixed set of clients. However, addressing the dynamic availability of clients -- a critical aspect of real-world scenarios -- remains underexplored. This study delves into optimizing client selection and client-to-edge associations in HFL under intermittent client participation so as to minimize overall system costs (i.e., delay and energy), while achieving fast model convergence. We unveil that achieving this goal involves solving a complex NP-hard problem. To tackle this, we propose a stagewise methodology that splits the solution into two stages, referred to as Plan A and Plan B. Plan A focuses on identifying long-term clients with high chance of participation in subsequent model training rounds. Plan B serves as a backup, selecting alternative clients when long-term clients are unavailable during model training rounds. This stagewise methodology offers a fresh perspective on client selection that can enhance both HFL and conventional FL via enabling low-overhead decision-making processes. Through evaluations on MNIST and CIFAR-10 datasets, we show that our methodology outperforms existing benchmarks in terms of model accuracy and system costs.

cs.LG↗

arXiv2024

Accelerating the Delivery of Data Services over Uncertain Mobile Crowdsensing Networks

The challenge of exchanging and processing of big data over mobile crowdsensing (MCS) networks calls for designing seamless data service provisioning mechanisms to enable utilization of resources of mobile devices/users for crowdsensing tasks. Although conventional onsite spot trading of resources based on real-time network conditions can facilitate data sharing, it often suffers from prohibitively long service provisioning delays and unavoidable trading failures due to requiring timely analysis of dynamic network environment. These limitations motivate us to investigate an integrated forward and spot trading mechanism (iFAST), which entails a novel hybrid data trading protocol with time efficiency, over uncertain MCS ecosystems. In iFAST, the sellers (i.e., mobile devices who can contribute data) can provide long-term or temporary sensing services to the buyers (i.e., sensing tasks). Specifically, it enables signing long-term contracts in advance of future transactions through a forward trading mode, via analyzing historical statistics of the network/market, for which the notion of overbooking is introduced and promoted. iFAST further encourages the buyers with unsatisfying service quality to recruit temporary sellers through a spot trading mode, considering the current network/market conditions. We analyze the fundamental blocks of iFAST and provide a case study to demonstrate its performance. Inspirations for future research directions of next-generation sensing and communication are summarized.

cs.DC↗

arXiv2023

Unleashing the Potential of Stage-Wise Decision-Making in Scheduling of Graph-Structured Tasks over Mobile Vehicular Clouds

To effectively process high volume of data across a fleet of dynamic and distributed vehicles, it is crucial to implement resource provisioning techniques that can provide reliable, cost-effective, and timely computing services. This article explores computation-intensive task scheduling over mobile vehicular clouds (MVCs). We use undirected weighted graphs (UWGs) to model both the execution of tasks and communication patterns among vehicles in an MVC. We then study reliable and timely scheduling of UWG tasks through a novel mechanism, operating on two complementary decision-making stages: Plan A and Plan B. Plan A entails a proactive decision-making approach, leveraging historical statistical data for the preemptive creation of an optimal mapping ($α$) between tasks and the MVC prior to practical task scheduling. In contrast, Plan B explores a real-time decision-making paradigm, functioning as a reliable contingency plan. It seeks a viable mapping ($β$) if $α$ encounters failures during task scheduling due to the unpredictable nature of the network. Furthermore, we provide an in-depth exploration of the procedural intricacies and key contributing factors that underpin the success of our mechanism. Additionally, we present a case study showcasing the superior performance on time efficiency and computation overhead. We further discuss a series of open directions for future research.

cs.DC↗

arXiv2023

Cloud-Magnetic Resonance Imaging System: In the Era of 6G and Artificial Intelligence

Magnetic Resonance Imaging (MRI) plays an important role in medical diagnosis, generating petabytes of image data annually in large hospitals. This voluminous data stream requires a significant amount of network bandwidth and extensive storage infrastructure. Additionally, local data processing demands substantial manpower and hardware investments. Data isolation across different healthcare institutions hinders cross-institutional collaboration in clinics and research. In this work, we anticipate an innovative MRI system and its four generations that integrate emerging distributed cloud computing, 6G bandwidth, edge computing, federated learning, and blockchain technology. This system is called Cloud-MRI, aiming at solving the problems of MRI data storage security, transmission speed, AI algorithm maintenance, hardware upgrading, and collaborative work. The workflow commences with the transformation of k-space raw data into the standardized Imaging Society for Magnetic Resonance in Medicine Raw Data (ISMRMRD) format. Then, the data are uploaded to the cloud or edge nodes for fast image reconstruction, neural network training, and automatic analysis. Then, the outcomes are seamlessly transmitted to clinics or research institutes for diagnosis and other services. The Cloud-MRI system will save the raw imaging data, reduce the risk of data loss, facilitate inter-institutional medical collaboration, and finally improve diagnostic accuracy and work efficiency.

eess.IV↗

arXiv2021

The Price Impact of Generalized Order Flow Imbalance

Order flow imbalance can explain short-term changes in stock price. This paper considers the change of non-minimum quotation units in real transactions, and proposes a generalized order flow imbalance construction method to improve Order Flow Imbalance (OFI) and Stationarized Order Flow Imbalance (log-OFI). Based on the high-frequency order book snapshot data, we conducted an empirical analysis of the CSI 500 constituent stocks. In order to facilitate the presentation, we selected 10 stocks for comparison. The two indicators after the improvement of the generalized order flow imbalance construction method both show a better ability to explain changes in stock prices. Especially Generalized Stationarized Order Flow Imbalance (log-GOFI), using a linear regression model, on the time scales of 30 seconds, 1 minute, and 5 minutes, the average R-squared out of sample compared with Order Flow Imbalance (OFI) 32.89%, 38.13% and 42.57%, respectively increased to 83.57%, 85.37% and 86.01%. In addition, we found that the interpretability of Generalized Stationarized Order Flow Imbalance (log-GOFI) showed stronger stability on all three time scales.

q-fin.TR↗