arXiv subjects

Hanwen Zhang

Publications and source records attributed to Hanwen Zhang.

At least 55 records · Page 3Linked to original sources

arXiv2025

A New One-Shot Federated Learning Framework for Medical Imaging Classification with Feature-Guided Rectified Flow and Knowledge Distillation

In multi-center scenarios, One-Shot Federated Learning (OSFL) has attracted increasing attention due to its low communication overhead, requiring only a single round of transmission. However, existing generative model-based OSFL methods suffer from low training efficiency and potential privacy leakage in the healthcare domain. Additionally, achieving convergence within a single round of model aggregation is challenging under non-Independent and Identically Distributed (non-IID) data. To address these challenges, in this paper a modified OSFL framework is proposed, in which a new Feature-Guided Rectified Flow Model (FG-RF) and Dual-Layer Knowledge Distillation (DLKD) aggregation method are developed. FG-RF on the client side accelerates generative modeling in medical imaging scenarios while preserving privacy by synthesizing feature-level images rather than pixel-level images. To handle non-IID distributions, DLKD enables the global student model to simultaneously mimic the output logits and align the intermediate-layer features of client-side teacher models during aggregation. Experimental results on three non-IID medical imaging datasets show that our new framework and method outperform multi-round federated learning approaches, achieving up to 21.73% improvement, and exceeds the baseline FedISCA by an average of 21.75%. Furthermore, our experiments demonstrate that feature-level synthetic images significantly reduce privacy leakage risks compared to pixel-level synthetic images. The code is available at https://github.com/LMIAPC/one-shot-fl-medical.

cs.CV↗

arXiv2025

High-Resolution Sustain Pedal Depth Estimation from Piano Audio Across Room Acoustics

Piano sustain pedal detection has previously been approached as a binary on/off classification task, limiting its application in real-world piano performance scenarios where pedal depth significantly influences musical expression. This paper presents a novel approach for high-resolution estimation that predicts continuous pedal depth values. We introduce a Transformer-based architecture that not only matches state-of-the-art performance on the traditional binary classification task but also achieves high accuracy in continuous pedal depth estimation. Furthermore, by estimating continuous values, our model provides musically meaningful predictions for sustain pedal usage, whereas baseline models struggle to capture such nuanced expressions with their binary detection approach. Additionally, this paper investigates the influence of room acoustics on sustain pedal estimation using a synthetic dataset that includes varied acoustic conditions. We train our model with different combinations of room settings and test it in an unseen new environment using a "leave-one-out" approach. Our findings show that the two baseline models and ours are not robust to unseen room conditions. Statistical analysis further confirms that reverberation influences model predictions and introduces an overestimation bias.

cs.SD↗

arXiv2025

Autoregressive Denoising Score Matching is a Good Video Anomaly Detector

Video anomaly detection (VAD) is an important computer vision problem. Thanks to the mode coverage capabilities of generative models, the likelihood-based paradigm is catching growing interest, as it can model normal distribution and detect out-of-distribution anomalies. However, these likelihood-based methods are blind to the anomalies located in local modes near the learned distribution. To handle these ``unseen" anomalies, we dive into three gaps uniquely existing in VAD regarding scene, motion and appearance. Specifically, we first build a noise-conditioned score transformer for denoising score matching. Then, we introduce a scene-dependent and motion-aware score function by embedding the scene condition of input sequences into our model and assigning motion weights based on the difference between key frames of input sequences. Next, to solve the problem of blindness in principle, we integrate unaffected visual information via a novel autoregressive denoising score matching mechanism for inference. Through autoregressively injecting intensifying Gaussian noise into the denoised data and estimating the corresponding score function, we compare the denoised data with the original data to get a difference and aggregate it with the score function for an enhanced appearance perception and accumulate the abnormal context. With all three gaps considered, we can compute a more comprehensive anomaly indicator. Experiments on three popular VAD benchmarks demonstrate the state-of-the-art performance of our method.

cs.CV↗

arXiv2025

URPlanner: A Universal Paradigm For Collision-Free Robotic Motion Planning Based on Deep Reinforcement Learning

Collision-free motion planning for redundant robot manipulators in complex environments is yet to be explored. Although recent advancements at the intersection of deep reinforcement learning (DRL) and robotics have highlighted its potential to handle versatile robotic tasks, current DRL-based collision-free motion planners for manipulators are highly costly, hindering their deployment and application. This is due to an overreliance on the minimum distance between the manipulator and obstacles, inadequate exploration and decision-making by DRL, and inefficient data acquisition and utilization. In this article, we propose URPlanner, a universal paradigm for collision-free robotic motion planning based on DRL. URPlanner offers several advantages over existing approaches: it is platform-agnostic, cost-effective in both training and deployment, and applicable to arbitrary manipulators without solving inverse kinematics. To achieve this, we first develop a parameterized task space and a universal obstacle avoidance reward that is independent of minimum distance. Second, we introduce an augmented policy exploration and evaluation algorithm that can be applied to various DRL algorithms to enhance their performance. Third, we propose an expert data diffusion strategy for efficient policy learning, which can produce a large-scale trajectory dataset from only a few expert demonstrations. Finally, the superiority of the proposed methods is comprehensively verified through experiments.

cs.RO↗

arXiv2025

CacheFL: Privacy-Preserving and Efficient Federated Cache Model Fine-Tuning for Vision-Language Models

Large pre-trained Vision-Language Models (VLMs), such as Contrastive Language-Image Pre-training (CLIP), have exhibited remarkable zero-shot performance across various image classification tasks. Fine-tuning these models on domain-specific datasets further enhances their effectiveness for downstream applications. However, fine-tuning in cloud environments raises significant concerns regarding data security and privacy. Federated Learning (FL) offers a decentralized solution by enabling model training across local clients without centralizing sensitive data, but the high communication and computation costs of transmitting full pre-trained models during training limit its scalability. Additionally, non-Independent and Identically Distributed (non-IID) data across local clients can negatively impact model convergence and performance. To address these challenges, we propose CacheFL, a novel federated learning method that replaces traditional full model fine-tuning with lightweight cache model fine-tuning. The cache model is initialized using a class-balanced dataset generated by a generative pre-trained model, effectively mitigating the impact of non-IID data. This cache model is then distributed to local clients for fine-tuning, and the updated parameters from each client are aggregated on the server and redistributed. With the updated cache model, the classification performance of CLIP is improved after just a few epochs. By limiting the training and communication to the cache model, CacheFL significantly reduces resource demands while ensuring data privacy and security. Extensive experiments conducted on ImageNet and 10 additional datasets demonstrate that CacheFL outperforms traditional approaches in terms of classification accuracy, resource efficiency, and privacy preservation.

cs.DC↗

arXiv2025

Constructing optimal Wannier functions via potential theory: isolated single band for matrix models

We present a rapidly convergent scheme for computing globally optimal Wannier functions of isolated single bands for matrix models in two dimensions. The scheme proceeds first by constructing provably exponentially localized Wannier functions directly from parallel transport (with simple analytically computable corrections) when topological obstructions are absent. We prove that the corresponding Wannier functions are real when the matrix model possesses time-reversal symmetry. When a band has a nonzero Berry curvature, the resulting Wannier function is not optimal, but it is transformed into the global optimum by a single gauge transformation that eliminates the divergence of the Berry connection. Complete analysis of the construction is presented, paving the way for further improvements and generalizations. The performance of the scheme is illustrated with several numerical examples.

math-ph↗

arXiv2025

HoP: Homeomorphic Polar Learning for Hard Constrained Optimization

Constrained optimization demands highly efficient solvers which promotes the development of learn-to-optimize (L2O) approaches. As a data-driven method, L2O leverages neural networks to efficiently produce approximate solutions. However, a significant challenge remains in ensuring both optimality and feasibility of neural networks' output. To tackle this issue, we introduce Homeomorphic Polar Learning (HoP) to solve the star-convex hard-constrained optimization by embedding homeomorphic mapping in neural networks. The bijective structure enables end-to-end training without extra penalty or correction. For performance evaluation, we evaluate HoP's performance across a variety of synthetic optimization tasks and real-world applications in wireless communications. In all cases, HoP achieves solutions closer to the optimum than existing L2O methods while strictly maintaining feasibility.

cs.LG↗

arXiv2025

CoMA: Compositional Human Motion Generation with Multi-modal Agents

3D human motion generation has seen substantial advancement in recent years. While state-of-the-art approaches have improved performance significantly, they still struggle with complex and detailed motions unseen in training data, largely due to the scarcity of motion datasets and the prohibitive cost of generating new training examples. To address these challenges, we introduce CoMA, an agent-based solution for complex human motion generation, editing, and comprehension. CoMA leverages multiple collaborative agents powered by large language and vision models, alongside a mask transformer-based motion generator featuring body part-specific encoders and codebooks for fine-grained control. Our framework enables generation of both short and long motion sequences with detailed instructions, text-guided motion editing, and self-correction for improved quality. Evaluations on the HumanML3D dataset demonstrate competitive performance against state-of-the-art methods. Additionally, we create a set of context-rich, compositional, and long text prompts, where user studies show our method significantly outperforms existing approaches.

cs.CV↗

arXiv2024

Optimal Bounds for Private Minimum Spanning Trees via Input Perturbation

We study the problem of privately releasing an approximate minimum spanning tree (MST). Given a graph $G = (V, E, \vec{W})$ where $V$ is a set of $n$ vertices, $E$ is a set of $m$ undirected edges, and $ \vec{W} \in \mathbb{R}^{|E|} $ is an edge-weight vector, our goal is to publish an approximate MST under edge-weight differential privacy, as introduced by Sealfon in PODS 2016, where $V$ and $E$ are considered public and the weight vector is private. Our neighboring relation is $\ell_\infty$-distance on weights: for a sensitivity parameter $Δ_\infty$, graphs $ G = (V, E, \vec{W}) $ and $ G' = (V, E, \vec{W}') $ are neighboring if $\|\vec{W}-\vec{W}'\|_\infty \leq Δ_\infty$. Existing private MST algorithms face a trade-off, sacrificing either computational efficiency or accuracy. We show that it is possible to get the best of both worlds: With a suitable random perturbation of the input that does not suffice to make the weight vector private, the result of any non-private MST algorithm will be private and achieves a state-of-the-art error guarantee. Furthermore, by establishing a connection to Private Top-k Selection [Steinke and Ullman, FOCS '17], we give the first privacy-utility trade-off lower bound for MST under approximate differential privacy, demonstrating that the error magnitude, $\tilde{O}(n^{3/2})$, is optimal up to logarithmic factors. That is, our approach matches the time complexity of any non-private MST algorithm and at the same time achieves optimal error. We complement our theoretical treatment with experiments that confirm the practicality of our approach.

cs.DS↗

arXiv2024

The Roles of Generative Artificial Intelligence in Internet of Electric Vehicles

With the advancements of generative artificial intelligence (GenAI) models, their capabilities are expanding significantly beyond content generation and the models are increasingly being used across diverse applications. Particularly, GenAI shows great potential in addressing challenges in the electric vehicle (EV) ecosystem ranging from charging management to cyber-attack prevention. In this paper, we specifically consider Internet of electric vehicles (IoEV) and we categorize GenAI for IoEV into four different layers namely, EV's battery layer, individual EV layer, smart grid layer, and security layer. We introduce various GenAI techniques used in each layer of IoEV applications. Subsequently, public datasets available for training the GenAI models are summarized. Finally, we provide recommendations for future directions. This survey not only categorizes the applications of GenAI in IoEV across different layers but also serves as a valuable resource for researchers and practitioners by highlighting the design and implementation challenges within each layer. Furthermore, it provides a roadmap for future research directions, enabling the development of more robust and efficient IoEV systems through the integration of advanced GenAI techniques.

cs.LG↗

arXiv2024

Faster Differentially Private Top-$k$ Selection: A Joint Exponential Mechanism with Pruning

We study the differentially private top-$k$ selection problem, aiming to identify a sequence of $k$ items with approximately the highest scores from $d$ items. Recent work by Gillenwater et al. (ICML '22) employs a direct sampling approach from the vast collection of $d^{\,Θ(k)}$ possible length-$k$ sequences, showing superior empirical accuracy compared to previous pure or approximate differentially private methods. Their algorithm has a time and space complexity of $\tilde{O}(dk)$. In this paper, we present an improved algorithm with time and space complexity $O(d + k^2 / ε\cdot \ln d)$, where $ε$ denotes the privacy parameter. Experimental results show that our algorithm runs orders of magnitude faster than their approach, while achieving similar empirical accuracy.

cs.CR↗

arXiv2024

Are Large Language Models Table-based Fact-Checkers?

Table-based Fact Verification (TFV) aims to extract the entailment relation between statements and structured tables. Existing TFV methods based on small-scaled models suffer from insufficient labeled data and weak zero-shot ability. Recently, the appearance of Large Language Models (LLMs) has gained lots of attraction in research fields. They have shown powerful zero-shot and in-context learning abilities on several NLP tasks, but their potential on TFV is still unknown. In this work, we implement a preliminary study about whether LLMs are table-based fact-checkers. In detail, we design diverse prompts to explore how the in-context learning can help LLMs in TFV, i.e., zero-shot and few-shot TFV capability. Besides, we carefully design and construct TFV instructions to study the performance gain brought by the instruction tuning of LLMs. Experimental results demonstrate that LLMs can achieve acceptable results on zero-shot and few-shot TFV with prompt engineering, while instruction-tuning can stimulate the TFV capability significantly. We also make some valuable findings about the format of zero-shot prompts and the number of in-context examples. Finally, we analyze some possible directions to promote the accuracy of TFV via LLMs, which is beneficial to further research of table reasoning.

cs.CL↗

arXiv2024

Model-based Deep Learning for Wireless Resource Allocation in RSMA Communications Systems

Rate-splitting multiple access (RSMA) has been proven as an effective communication scheme for 5G and beyond. However, current approaches to RSMA resource management require complicated iterative algorithms, which cannot meet the stringent latency requirement by users with limited resources. Recently, data-driven methods are explored to alleviate this issue. However, they suffer from poor generalizability and scarce training data to achieve satisfactory performance. In this paper, we propose a fractional programming (FP) based deep unfolding (DU) approach to address resource allocation problem for a weighted sum rate optimization in RSMA. By carefully designing the penalty function, we couple the variable update with projected gradient descent algorithm (PGD). Following the structure of PGD, we embed a few learnable parameters in each layer of the DU network. Through extensive simulation, we have shown that the proposed model-based neural networks can yield similar results compared to the traditional optimization algorithm for RSMA resource management but with much lower computational complexity, less training data, and higher resilience to out-of-distribution (OOD) data.

cs.IT↗

arXiv2024

Model-based Deep Learning for QoS-Aware Rate-Splitting Multiple Access Wireless Systems

Next generation communications demand for better spectrum management, lower latency, and guaranteed quality-of-service (QoS). Recently, Artificial intelligence (AI) has been widely introduced to advance these aspects in next generation wireless systems. However, such AI applications suffer from limited training data, low robustness, and poor generalization capabilities. To address these issues, a model-driven deep unfolding (DU) algorithm is introduced in this paper to bridge the gap between traditional model-driven communication algorithms and data-driven deep learning. Focusing on the QoS-aware rate-splitting multiple access (RSMA) resource allocation problem in multi-user communications, a conventional fractional programming (FP) algorithm is first applied as a benchmark. The solution is then refined by the application of projection gradient descent (PGD). DU is employed to further speed up convergence procedure, hence improving the efficiency of PGD. Moreover, the feasibility of results is guaranteed by designing a low-complexity projection based on scale factors, plus adding violation control mechanisms into the loss function that minimizes error rates. Finally, we provide a detailed analysis of the computational complexity and analysis design of the proposed DU algorithm. Extensive simulations are conducted and the results demonstrate that the proposed DU algorithm can reach the optimal communication efficiency with a mere $0.024\%$ violation rate for 4 layers DU. The DU algorithm also exhibits robustness in out-of-distribution tests and can be effectively trained with as few as 50 samples.

cs.NI↗

arXiv2024

Combinatorial Correlation Clustering

Correlation Clustering is a classic clustering objective arising in numerous machine learning and data mining applications. Given a graph $G=(V,E)$, the goal is to partition the vertex set into clusters so as to minimize the number of edges between clusters plus the number of edges missing within clusters. The problem is APX-hard and the best known polynomial time approximation factor is 1.73 by Cohen-Addad, Lee, Li, and Newman [FOCS'23]. They use an LP with $|V|^{1/ε^{Θ(1)}}$ variables for some small $ε$. However, due to the practical relevance of correlation clustering, there has also been great interest in getting more efficient sequential and parallel algorithms. The classic combinatorial \emph{pivot} algorithm of Ailon, Charikar and Newman [JACM'08] provides a 3-approximation in linear time. Like most other algorithms discussed here, this uses randomization. Recently, Behnezhad, Charikar, Ma and Tan [FOCS'22] presented a $3+ε$-approximate solution for solving problem in a constant number of rounds in the Massively Parallel Computation (MPC) setting. Very recently, Cao, Huang, Su [SODA'24] provided a 2.4-approximation in a polylogarithmic number of rounds in the MPC model and in $\tilde{O} (|E|^{1.5})$ time in the classic sequential setting. They asked whether it is possible to get a better than 3-approximation in near-linear time? We resolve this problem with an efficient combinatorial algorithm providing a drastically better approximation factor. It achieves a $\sim 2-2/13 < 1.847$-approximation in sub-linear ($\tilde O(|V|)$) sequential time or in sub-linear ($\tilde O(|V|)$) space in the streaming setting. In the MPC model, we give an algorithm using only a constant number of rounds that achieves a $\sim 2-1/8 < 1.876$-approximation.

cs.DS↗

arXiv2024

Exploring Communication Technologies, Standards, and Challenges in Electrified Vehicle Charging

As public awareness of environmental protection continues to grow, the trend of integrating more electric vehicles (EVs) into the transportation sector is rising. Unlike conventional internal combustion engine (ICE) vehicles, EVs can minimize carbon emissions and potentially achieve autonomous driving. However, several obstacles hinder the widespread adoption of EVs, such as their constrained driving range and the extended time required for charging. One alternative solution to address these challenges is implementing dynamic wireless power transfer (DWPT), charging EVs in motion on the road. Moreover, charging stations with static wireless power transfer (SWPT) infrastructure can replace existing gas stations, enabling users to charge EVs in parking lots or at home. This paper surveys the communication infrastructure for static and dynamic wireless charging in electric vehicles. It encompasses all communication aspects involved in the wireless charging process. The architecture and communication requirements for static and dynamic wireless charging are presented separately. Additionally, a comprehensive comparison of existing communication standards is provided. The communication with the grid is also explored in detail. The survey gives attention to security and privacy issues arising during communications. In summary, the paper addresses the challenges and outlines upcoming trends in communication for EV wireless charging.

cs.NI↗

arXiv2024

MPCPA: Multi-Center Privacy Computing with Predictions Aggregation based on Denoising Diffusion Probabilistic Model

Privacy-preserving computing is crucial for multi-center machine learning in many applications such as healthcare and finance. In this paper a Multi-center Privacy Computing framework with Predictions Aggregation (MPCPA) based on denoising diffusion probabilistic model (DDPM) is proposed, in which conditional diffusion model training, DDPM data generation, a classifier, and strategy of prediction aggregation are included. Compared to federated learning, this framework necessitates fewer communications and leverages high-quality generated data to support robust privacy computing. Experimental validation across multiple datasets demonstrates that the proposed framework outperforms classic federated learning and approaches the performance of centralized learning with original data. Moreover, our approach demonstrates robust security, effectively addressing challenges such as image memorization and membership inference attacks. Our experiments underscore the efficacy of the proposed framework in the realm of privacy computing, with the code set to be released soon.

cs.DC↗

arXiv2024

Viscoelastic mechanics of tidally induced lake drainage in the Amery grounding zone

Drainage of supraglacial lakes to the ice-sheet bed can occur when a hydrofracture propagates downward, driven by the weight of the water in the lake. For supraglacial lakes in the grounding zones of Antarctic glaciers, the mechanics of drainage is complicated by their proximity to the grounding line. Recently, a series of supraglacial lake-drainage events through hydrofractures was observed in the Amery Ice Shelf grounding zone, East Antarctica. The lake depth at drainage varied considerably between events, raising questions about the mechanisms that induce hydrofracture, even at low lake depths. Here we use a modelling approach to investigate the contribution of tidally driven flexure to hydrofracture propagation. We model the viscoelastic response of a marine ice sheet to tides, the stresses that are induced, and the contribution of tidal stresses to hydrofracture propagation. Our results show that ocean tides and lake-water pressure together control supraglacial lake drainage through hydrofractures in the grounding zone. We give a model-based criterion that predicts supraglacial lake drainage as a function of daily maximum tidal amplitude and lake depth. Our model-based criterion agrees with remotely sensed data, indicating the importance of tidal flexure to processes associated with hydrofracturing such as supraglacial lake drainage, rifting and calving.

physics.geo-ph↗