arXiv subjects

Su Zhang

Publications and source records attributed to Su Zhang.

At least 19 recordsLinked to original sources

arXiv2026

Low-cost IoT-Based Rainfall Monitoring with Web-Based Data Access

Rainfall measurement with high spatial and temporal resolution is critical for flood forecasting, drought mitigation, and disaster preparedness. Rainfall patterns are highly variable, both geographically and over time. This variability presents a significant challenge for monitoring, as rain gauges can accurately capture temporal patterns only at a single location. Furthermore, the high cost of commercial instruments restricts their widespread deployment, and rain gauge networks often fail to adequately capture the spatial heterogeneity of precipitation patterns. To address these limitations, this study introduces a low-cost IoT-based rainfall monitoring system developed upon the Low-cost Efficient Wireless Intelligent Sensor (LEWIS) platform. Four rainfall sensors were designed, developed, and deployed at different locations across the semi-arid region of the United States, in the State of New Mexico, to capture localized precipitation variability. Each sensor node integrates a rainfall detection module with an LTE-enabled microcontroller and is powered by a compact solar-battery system, ensuring autonomous and self-sufficient operation. Real-time precipitation data are transmitted to a cloud server for continuous access, visualization, and integration with early-warning frameworks. The results demonstrate that IoT-based rainfall monitoring can achieve reliable accuracy at a fraction of the cost of conventional gauges, while supporting dense deployment for microscale precipitation analysis. Comparative validation with model-based precipitation data and in situ observations shows strong agreement in the detection and timing of recorded precipitation events, highlighting the system potential for early warning, disaster risk reduction, and bias correction of remotely sensed precipitation products by filling observational gaps in under-instrumented semi-arid areas.

eess.SP↗

arXiv2026

FedAttr: Towards Privacy-preserving Client-Level Attribution in Federated LLM Fine-tuning

Watermark radioactivity testing type of methods can detect whether a model was trained on watermarked documents, and have become key tools for protecting data ownership in the fine-tuning of large language models (LLMs). Existing works have proved their effectiveness in centralized LLM fine-tuning. However, this type of method faces several challenges and remains underexplored in federated learning (FL), a widely-applied paradigm for fine-tuning LLMs collaboratively on private data across different users. FL mainly ensures privacy through secure aggregation (SA), which allows the server to aggregate updates while keeping clients' updates private. This mechanism preserves privacy but makes it difficult to identify which client trained on watermarked documents. In this work, we propose FedAttr, a new client-level attribution protocol for FL. FedAttr identifies which clients trained on watermarked data via a paired-subset-difference mechanism, while preserving the privacy guarantees of SA and FL performance. FedAttr proceeds in three steps: (i) estimate each client's update by differencing two SA queries, (ii) score the estimate with the watermark detector via differential scoring, and (iii) combine scores across rounds via Stouffer method. We theoretically show that FedAttr produces an unbiased estimator of each client's update with bounded mutual information leakage (i.e., $O(d^*/N)$ per-round update). Moreover, FedAttr empirically achieves 100% TPR and 0% FPR, outperforming all baselines by at least 44.4% in TPR or 19.1% in FPR, with only 6.3% overhead relative to FL training time. Ablation studies confirm that FedAttr is robust to protocol parameters and configurations.

cs.CR↗

arXiv2026

A Recursive Polynomial Chaos Evolution Method for Stochastic Differential Equations

Numerical simulation of stochastic differential equations over long time intervals poses significant computational challenges. In this paper, we propose a novel recursive polynomial chaos evolution method that achieves model reduction without sampling by exploiting the Markov property to maintain a fixed low-dimensional representation throughout the time evolution. At each time step, we construct orthogonal polynomial bases adapted to the current probability measure, and project the one-step-ahead solution onto this new basis together with the new Brownian increments. This dynamic updating strategy effectively reduces the dimension of the random variables during long-time evolution. Under appropriate assumptions, we prove the convergence of the method, specifically that the distributions generated by the method preserve convergence in the Wasserstein-1 distance. We present numerical results demonstrating that the method can accurately capture complex dynamical behaviors with high accuracy and low computational cost.

math.NA↗

arXiv2026

Federated Multi Agent Deep Learning and Neural Networks for Advanced Distributed Sensing in Wireless Networks

Multi-agent deep learning (MADL), including multi-agent deep reinforcement learning (MADRL), distributed/federated training, and graph-structured neural networks, is becoming a unifying framework for decision-making and inference in wireless systems where sensing, communication, and computing are tightly coupled. Recent 5G-Advanced and 6G visions strengthen this coupling through integrated sensing and communication, edge intelligence, open programmable RAN, and non-terrestrial/UAV networking, which create decentralized, partially observed, time-varying, and resource-constrained control problems. This survey synthesizes the state of the art, with emphasis on 2021-2025 research, on MADL for distributed sensing and wireless communications. We present a task-driven taxonomy across (i) learning formulations (Markov games, Dec-POMDPs, CTDE), (ii) neural architectures (GNN-based radio resource management, attention-based policies, hierarchical learning, and over-the-air aggregation), (iii) advanced techniques (federated reinforcement learning, communication-efficient federated deep RL, and serverless edge learning orchestration), and (iv) application domains (MEC offloading with slicing, UAV-enabled heterogeneous networks with power-domain NOMA, intrusion detection in sensor networks, and ISAC-driven perceptive mobile networks). We also provide comparative tables of algorithms, training topologies, and system-level trade-offs in latency, spectral efficiency, energy, privacy, and robustness. Finally, we identify open issues including scalability, non-stationarity, security against poisoning and backdoors, communication overhead, and real-time safety, and outline research directions toward 6G-native sense-communicate-compute-learn systems.

cs.LG↗

arXiv2025

Trade-off in Estimating the Number of Byzantine Clients in Federated Learning

Federated learning has attracted increasing attention at recent large-scale optimization and machine learning research and applications, but is also vulnerable to Byzantine clients that can send any erroneous signals. Robust aggregators are commonly used to resist Byzantine clients. This usually requires to estimate the unknown number $f$ of Byzantine clients, and thus accordingly select the aggregators with proper degree of robustness (i.e., the maximum number $\hat{f}$ of Byzantine clients allowed by the aggregator). Such an estimation should have important effect on the performance, which has not been systematically studied to our knowledge. This work will fill in the gap by theoretically analyzing the worst-case error of aggregators as well as its induced federated learning algorithm for any cases of $\hat{f}$ and $f$. Specifically, we will show that underestimation ($\hat{f}<f$) can lead to arbitrarily poor performance for both aggregators and federated learning. For non-underestimation ($\hat{f}\ge f$), we have proved optimal lower and upper bounds of the same order on the errors of both aggregators and federated learning. All these optimal bounds are proportional to $\hat{f}/(n-f-\hat{f})$ with $n$ clients, which monotonically increases with larger $\hat{f}$. This indicates a fundamental trade-off: while an aggregator with a larger robustness degree $\hat{f}$ can solve federated learning problems of wider range $f\in [0,\hat{f}]$, the performance can deteriorate when there are actually fewer or even no Byzantine clients (i.e., $f\in [0,\hat{f})$).

cs.LG↗

arXiv2025

RubikSQL: Lifelong Learning Agentic Knowledge Base as an Industrial NL2SQL System

We present RubikSQL, a novel NL2SQL system designed to address key challenges in real-world enterprise-level NL2SQL, such as implicit intents and domain-specific terminology. RubikSQL frames NL2SQL as a lifelong learning task, demanding both Knowledge Base (KB) maintenance and SQL generation. RubikSQL systematically builds and refines its KB through techniques including database profiling, structured information extraction, agentic rule mining, and Chain-of-Thought (CoT)-enhanced SQL profiling. RubikSQL then employs a multi-agent workflow to leverage this curated KB, generating accurate SQLs. RubikSQL achieves SOTA performance on both the KaggleDBQA and BIRD Mini-Dev datasets. Finally, we release the RubikBench benchmark, a new benchmark specifically designed to capture vital traits of industrial NL2SQL scenarios, providing a valuable resource for future research.

cs.DB↗

arXiv2025

Shadow in the Cache: Unveiling and Mitigating Privacy Risks of KV-cache in LLM Inference

The Key-Value (KV) cache, which stores intermediate attention computations (Key and Value pairs) to avoid redundant calculations, is a fundamental mechanism for accelerating Large Language Model (LLM) inference. However, this efficiency optimization introduces significant yet underexplored privacy risks. This paper provides the first comprehensive analysis of these vulnerabilities, demonstrating that an attacker can reconstruct sensitive user inputs directly from the KV-cache. We design and implement three distinct attack vectors: a direct Inversion Attack, a more broadly applicable and potent Collision Attack, and a semantic-based Injection Attack. These methods demonstrate the practicality and severity of KV-cache privacy leakage issues. To mitigate this, we propose KV-Cloak, a novel, lightweight, and efficient defense mechanism. KV-Cloak uses a reversible matrix-based obfuscation scheme, combined with operator fusion, to secure the KV-cache. Our extensive experiments show that KV-Cloak effectively thwarts all proposed attacks, reducing reconstruction quality to random noise. Crucially, it achieves this robust security with virtually no degradation in model accuracy and minimal performance overhead, offering a practical solution for trustworthy LLM deployment.

cs.CR↗

arXiv2025

Learning event-triggered controllers for linear parameter-varying systems from data

Nonlinear dynamical behaviours in engineering applications can be approximated by linear-parameter varying (LPV) representations, but obtaining precise model knowledge to develop a control algorithm is difficult in practice. In this paper, we develop the data-driven control strategies for event-triggered LPV systems with stability verifications. First, we provide the theoretical analysis of ${\theta}$-persistence of excitation for LPV systems, which leads to the feasible data-based representations. Then, in terms of the available perturbed data, we derive the stability certificates for event-triggered LPV systems with the aid of Petersen's lemma in the sense of robust control, resulting in the computationally tractable semidefinite programmings, the feasible solutions of which yields the optimal gain schedulings. Besides, we generalize the data-driven eventtriggered LPV control methods to the scenario of reference trajectory tracking, and discuss the robust tracking stability accordingly. Finally, we verify the effectiveness of our theoretical derivations by numerical simulations.

eess.SY↗

arXiv2024

A Unified Intracellular pH Landscape with SITE-pHorin: a Quantum-Entanglement-Enhanced pH Probe

An accurate map of intracellular organelle pH is crucial for comprehending cellular metabolism and organellar functions. However, a unified intracellular pH spectrum using a single probe is still lack. Here, we developed a novel quantum entanglement-enhanced pH-sensitive probe called SITE-pHorin, which featured a wide pH-sensitive range and ratiometric quantitative measurement capabilities. Subsequently, we measured the pH of various organelles and their sub-compartments, including mitochondrial sub-spaces, Golgi stacks, endoplasmic reticulum, lysosomes, peroxisomes, and endosomes in COS-7 cells. For the long-standing debate on mitochondrial compartments pH, we measured the pH of mitochondrial cristae as 6.60 \pm 0.40, the pH of mitochondrial intermembrane space as 6.95 \pm 0.30, and two populations of mitochondrial matrix pH at approximately 7.20 \pm 0.27 and 7.50 \pm 0.16, respectively. Notably, the lysosome pH exhibited a single, narrow Gaussian distribution centered at 4.79 \pm 0.17. Furthermore, quantum chemistry computations revealed that both the deprotonation of the residue Y182 and the discrete curvature of deformed benzene ring in chromophore are both necessary for the quantum entanglement mechanism of SITE-pHorin. Intriguingly, our findings reveal an accurate pH gradient (0.6-0.9 pH unit) between mitochondrial cristae and matrix, suggesting prior knowledge about \Delta pH (0.4-0.6) and mitochondrial proton motive force (pmf) are underestimated.

q-bio.QM↗

arXiv2024

FlamePINN-1D: Physics-informed neural networks to solve forward and inverse problems of 1D laminar flames

Given the existence of various forward and inverse problems in combustion studies and applications that necessitate distinct methods for resolution, a framework to solve them in a unified way is critically needed. A promising approach is the integration of machine learning methods with governing equations of combustion systems, which exhibits superior generality and few-shot learning ability compared to purely data-driven methods. In this work, the FlamePINN-1D framework is proposed to solve the forward and inverse problems of 1D laminar flames based on physics-informed neural networks. Three cases with increasing complexity have been tested: Case 1 are freely-propagating premixed (FPP) flames with simplified physical models, while Case 2 and Case 3 are FPP and counterflow premixed (CFP) flames with detailed models, respectively. For forward problems, FlamePINN-1D aims to solve the flame fields and infer the unknown eigenvalues (such as laminar flame speeds) under the constraints of governing equations and boundary conditions. For inverse problems, FlamePINN-1D aims to reconstruct the continuous fields and infer the unknown parameters (such as transport and chemical kinetics parameters) from noisy sparse observations of the flame. Our results strongly validate these capabilities of FlamePINN-1D across various flames and working conditions. Compared to traditional methods, FlamePINN-1D is differentiable and mesh-free, exhibits no discretization errors, and is easier to implement for inverse problems. The inverse problem results also indicate the possibility of optimizing chemical mechanisms from measurements of laboratory 1D flames. Furthermore, some proposed strategies, such as hard constraints and thin-layer normalization, are proven to be essential for the robust learning of FlamePINN-1D. The code for this paper is partially available at https://github.com/CAME-THU/FlamePINN-1D.

cs.LG↗

arXiv2023

Bicoptor 2.0: Addressing Challenges in Probabilistic Truncation for Enhanced Privacy-Preserving Machine Learning

This paper primarily focuses on analyzing the problems and proposing solutions for the probabilistic truncation protocol in existing PPML works from the perspectives of accuracy and efficiency. In terms of accuracy, we reveal that precision selections recommended in some of the existing works are incorrect. We conduct a thorough analysis of their open-source code and find that their errors were mainly due to simplified implementation, more specifically, fixed numbers are used instead of random numbers in probabilistic truncation protocols. Based on this, we provide a detailed theoretical analysis to validate our views. We propose a solution and a precision selection guideline for future works. Regarding efficiency, we identify limitations in the state-of-the-art comparison protocol, Bicoptor's (S\&P 2023) DReLU protocol, which relies on the probabilistic truncation protocol and is heavily constrained by the security parameter to avoid errors, significantly impacting the protocol's performance. To address these challenges, we introduce the first non-interactive deterministic truncation protocol, replacing the original probabilistic truncation protocol. Additionally, we design a non-interactive modulo switch protocol to enhance the protocol's security. Finally, we provide a guideline to reduce computational and communication overhead by using only a portion of the bits of the input, i.e., the key bits, for DReLU operations based on different model parameters. With the help of key bits, the performance of our DReLU protocol is further improved. We evaluate the performance of our protocols on three GPU servers, and achieve a 10x improvement in DReLU protocol, and a 6x improvement in the ReLU protocol over the state-of-the-art work Piranha-Falcon (USENIX Sec 22). Overall, the performance of our end-to-end (E2E) privacy-preserving machine learning (PPML) inference is improved by 3-4 times.

cs.CR↗

arXiv2023

MASA-TCN: Multi-anchor Space-aware Temporal Convolutional Neural Networks for Continuous and Discrete EEG Emotion Recognition

Emotion recognition using electroencephalogram (EEG) mainly has two scenarios: classification of the discrete labels and regression of the continuously tagged labels. Although many algorithms were proposed for classification tasks, there are only a few methods for regression tasks. For emotion regression, the label is continuous in time. A natural method is to learn the temporal dynamic patterns. In previous studies, long short-term memory (LSTM) and temporal convolutional neural networks (TCN) were utilized to learn the temporal contextual information from feature vectors of EEG. However, the spatial patterns of EEG were not effectively extracted. To enable the spatial learning ability of TCN towards better regression and classification performances, we propose a novel unified model, named MASA-TCN, for EEG emotion regression and classification tasks. The space-aware temporal layer enables TCN to additionally learn from spatial relations among EEG electrodes. Besides, a novel multi-anchor block with attentive fusion is proposed to learn dynamic temporal dependencies. Experiments on two publicly available datasets show MASA-TCN achieves higher results than the state-of-the-art methods for both EEG emotion regression and classification tasks. The code is available at https://github.com/yi-ding-cs/MASA-TCN.

cs.LG↗

arXiv2023

Multimodal Continuous Emotion Recognition: A Technical Report for ABAW5

We used two multimodal models for continuous valence-arousal recognition using visual, audio, and linguistic information. The first model is the same as we used in ABAW2 and ABAW3, which employs the leader-follower attention. The second model has the same architecture for spatial and temporal encoding. As for the fusion block, it employs a compact and straightforward channel attention, borrowed from the End2You toolkit. Unlike our previous attempts that use Vggish feature directly as the audio feature, this time we feed the pre-trained VGG model using logmel-spectrogram and finetune it during the training. To make full use of the data and alleviate over-fitting, cross-validation is carried out. The code is available at https://github.com/sucv/ABAW3.

cs.MM↗

arXiv2022

Continuous Emotion Recognition using Visual-audio-linguistic information: A Technical Report for ABAW3

We propose a cross-modal co-attention model for continuous emotion recognition using visual-audio-linguistic information. The model consists of four blocks. The visual, audio, and linguistic blocks are used to learn the spatial-temporal features of the multi-modal input. A co-attention block is designed to fuse the learned features with the multi-head co-attention mechanism. The visual encoding from the visual block is concatenated with the attention feature to emphasize the visual information. To make full use of the data and alleviate over-fitting, cross-validation is carried out on the training and validation set. The concordance correlation coefficient (CCC) centering is used to merge the results from each fold. The achieved CCC on the test set is $0.520$ for valence and $0.602$ for arousal, which significantly outperforms the baseline method with the corresponding CCC of 0.180 and 0.170 for valence and arousal, respectively. The code is available at https://github.com/sucv/ABAW3.

cs.MM↗

arXiv2021

Continuous Emotion Recognition with Audio-visual Leader-follower Attentive Fusion

We propose an audio-visual spatial-temporal deep neural network with: (1) a visual block containing a pretrained 2D-CNN followed by a temporal convolutional network (TCN); (2) an aural block containing several parallel TCNs; and (3) a leader-follower attentive fusion block combining the audio-visual information. The TCN with large history coverage enables our model to exploit spatial-temporal information within a much larger window length (i.e., 300) than that from the baseline and state-of-the-art methods (i.e., 36 or 48). The fusion block emphasizes the visual modality while exploits the noisy aural modality using the inter-modality attention mechanism. To make full use of the data and alleviate over-fitting, cross-validation is carried out on the training and validation set. The concordance correlation coefficient (CCC) centering is used to merge the results from each fold. On the test (validation) set of the Aff-Wild2 database, the achieved CCC is 0.463 (0.469) for valence and 0.492 (0.649) for arousal, which significantly outperforms the baseline method with the corresponding CCC of 0.200 (0.210) and 0.190 (0.230) for valence and arousal, respectively. The code is available at https://github.com/sucv/ABAW2.

cs.CV↗

arXiv2021

TSception: Capturing Temporal Dynamics and Spatial Asymmetry from EEG for Emotion Recognition

The high temporal resolution and the asymmetric spatial activations are essential attributes of electroencephalogram (EEG) underlying emotional processes in the brain. To learn the temporal dynamics and spatial asymmetry of EEG towards accurate and generalized emotion recognition, we propose TSception, a multi-scale convolutional neural network that can classify emotions from EEG. TSception consists of dynamic temporal, asymmetric spatial, and high-level fusion layers, which learn discriminative representations in the time and channel dimensions simultaneously. The dynamic temporal layer consists of multi-scale 1D convolutional kernels whose lengths are related to the sampling rate of EEG, which learns the dynamic temporal and frequency representations of EEG. The asymmetric spatial layer takes advantage of the asymmetric EEG patterns for emotion, learning the discriminative global and hemisphere representations. The learned spatial representations will be fused by a high-level fusion layer. Using more generalized cross-validation settings, the proposed method is evaluated on two publicly available datasets DEAP and MAHNOB-HCI. The performance of the proposed network is compared with prior reported methods such as SVM, KNN, FBFgMDM, FBTSC, Unsupervised learning, DeepConvNet, ShallowConvNet, and EEGNet. TSception achieves higher classification accuracies and F1 scores than other methods in most of the experiments. The codes are available at https://github.com/yi-ding-cs/TSception

cs.LG↗

arXiv2019

Assessing Fashion Recommendations: A Multifaceted Offline Evaluation Approach

Fashion is a unique domain for developing recommender systems (RS). Personalization is critical to fashion users. As a result, highly accurate recommendations are not sufficient unless they are also specific to users. Moreover, fashion data is characterized by a large majority of new users, so a recommendation strategy that performs well only for users with prior interaction history is a poor fit to the fashion problem. Critical to addressing these issues in fashion recommendation is an evaluation strategy that: 1) includes multiple metrics that are relevant to fashion, and 2) is performed within segments of users with different interaction histories. Here, we present our multifaceted offline strategy for evaluating fashion RS. Using our proposed evaluation methodology, we compare the performance of three different algorithms, a most popular (MP) items strategy, a collaborative filtering (CF) strategy, and a content-based (CB) strategy. We demonstrate that only by considering the performance of these algorithms across multiple metrics and user segments can we determine the extent to which each algorithm is likely to fulfill fashion users' needs.

cs.IR↗

arXiv2019

Apply SOA Paradigms in Cyber-Physical System to Enhance Interoperability: State-of-the-Art Review

The Cyber-Physical System (CPS) is considered to be the next generation of intelligent industrial automation systems that integrate computing, communication and control technologies. In CPS, the interoperability requirements between devices and processes are growing rapidly. However, the diversity and heterogeneity of technologies and standards between information technology (IT) and operational technology (OT) in industrial systems pose significant challenges. Service-Oriented Architecture (SOA) provides a software design pattern in which software components are connected only through messaging, which is consistent with the fundamentals of modularity and communication in distributed automation systems. SOA is considered to be a key technology for improving interoperability in next-generation industrial automation systems. This paper introduces the latest research advances in applying SOA paradigms to CPS to improve interoperability, including: Open Platform Communications Unified Architecture (OPC UA) and related extensions, implementing SOA at the manufacturing device level, integrating legacy systems in SOA-based automation systems, Plug-and-Play in SOA-based automation systems, dynamic configuration in SOA-based automation systems and the combination of SOA with Multi-Agent Systems and Evolvable Production Systems.

cs.SE↗