arXiv subjects

Dongyang Li

Publications and source records attributed to Dongyang Li.

At least 37 records · Page 2Linked to original sources

arXiv2024

Multi-Objective Communication Optimization for Temporal Continuity in Dynamic Vehicular Networks

Vehicular Ad-hoc Networks (VANETs) operate in highly dynamic environments characterized by high mobility, time-varying channel conditions, and frequent network disruptions. Addressing these challenges, this paper presents a novel temporal-aware multi-objective robust optimization framework, which for the first time formally incorporates temporal continuity into the optimization of dynamic multi-hop VANETs. The proposed framework simultaneously optimizes communication delay, throughput, and reliability, ensuring stable and consistent communication paths under rapidly changing conditions. A robust optimization model is formulated to mitigate performance degradation caused by uncertainties in vehicular density and channel fluctuations. To solve the optimization problem, an enhanced Non-dominated Sorting Genetic Algorithm II (NSGA-II) is developed, integrating dynamic encoding, elite inheritance, and adaptive constraint handling to efficiently balance trade-offs among conflicting objectives. Simulation results demonstrate that the proposed framework achieves significant improvements in reliability, delay reduction, and throughput enhancement, while temporal continuity effectively stabilizes communication paths over time. This work provides a pioneering and comprehensive solution for optimizing VANET communication, offering critical insights for robust and efficient strategies in intelligent transportation systems.

cs.NE↗

arXiv2024

Adaptive Genetic Selection based Pinning Control with Asymmetric Coupling for Multi-Network Heterogeneous Vehicular Systems

To alleviate computational load on RSUs and cloud platforms, reduce communication bandwidth requirements, and provide a more stable vehicular network service, this paper proposes an optimized pinning control approach for heterogeneous multi-network vehicular ad-hoc networks (VANETs). In such networks, vehicles participate in multiple task-specific networks with asymmetric coupling and dynamic topologies. We first establish a rigorous theoretical foundation by proving the stability of pinning control strategies under both single and multi-network conditions, deriving sufficient stability conditions using Lyapunov theory and linear matrix inequalities (LMIs). Building on this theoretical groundwork, we propose an adaptive genetic algorithm tailored to select optimal pinning nodes, effectively balancing LMI constraints while prioritizing overlapping nodes to enhance control efficiency. Extensive simulations across various network scales demonstrate that our approach achieves rapid consensus with a reduced number of control nodes, particularly when leveraging network overlaps. This work provides a comprehensive solution for efficient control node selection in complex vehicular networks, offering practical implications for deploying large-scale intelligent transportation systems.

cs.AI↗

arXiv2024

Blockchain-Based Multi-Path Mobile Access Point Selection for Secure 5G VANETs

This letter presents a blockchain-based multi-path mobile access point (MAP) selection strategy for secure 5G vehicular ad-hoc networks (VANETs). The proposed method leverages blockchain technology for decentralized, transparent, and secure MAP selection, while the multi-path transmission strategy enhances network reliability and reduces communication delays. A trust-based attack detection mechanism is integrated to ensure network security. Simulation results demonstrate that the proposed algorithm reduces both handover frequency and average communication delay by over 80%, and successfully identifies and excludes more than 95% of Sybil nodes, ensuring reliable and secure communication in highly dynamic vehicular environments.

cs.CR↗

arXiv2024

RealMind: Advancing Visual Decoding and Language Interaction via EEG Signals

Decoding visual stimuli from neural recordings is a critical challenge in the development of brain-computer interfaces (BCIs). Although recent EEG-based decoding approaches have made progress in tasks such as visual classification, retrieval, and reconstruction, they remain constrained by unstable representation learning and a lack of interpretability. This gap highlights the need for more efficient representation learning and the integration of effective language interaction to enhance both understanding and practical usability in visual decoding tasks.To address this limitation, we introduce RealMind, a novel EEG-based framework designed to handle a diverse range of downstream tasks. Specifically, RealMind leverages both semantic and geometric consistency learning to enhance feature representation and improve alignment across tasks. Notably, beyond excelling in traditional tasks, our framework marks the first attempt at visual captioning from EEG data through vision-language model (VLM). It achieves a Top-1 decoding accuracy of 27.58% in a 200-class zero-shot retrieval task and a BLEU-1 score of 26.59% in a 200-class zero-shot captioning task. Overall, RealMind provides a comprehensive multitask EEG decoding framework, establishing a foundational approach for EEG-based visual decoding in real-world applications.

cs.HC↗

arXiv2024

Random space-time sampling and reconstruction of sparse bandlimited graph diffusion field

In this work, we investigate the sampling and reconstruction of spectrally $s$-sparse bandlimited graph signals governed by heat diffusion processes. We propose a random space-time sampling regime, referred to as {randomized} dynamical sampling, where a small subset of space-time nodes is randomly selected at each time step based on a probability distribution. To analyze the recovery problem, we establish a rigorous mathematical framework by introducing the parameter \textit{the dynamic spectral graph weighted coherence}. This key parameter governs the number of space-time samples needed for stable recovery and extends the idea of variable density sampling to the context of dynamical systems. By optimizing the sampling probability distribution, we show that as few as $\mathcal{O}(s \log(k))$ space-time samples are sufficient for accurate reconstruction in optimal scenarios, where $k$ denotes the bandwidth of the signal. Our framework encompasses both static and dynamic cases, demonstrating a reduction in the number of spatial samples needed at each time step by exploiting temporal correlations. Furthermore, we provide a computationally efficient and robust algorithm for signal reconstruction. Numerical experiments validate our theoretical results and illustrate the practical efficacy of our proposed methods.

math.NA↗

arXiv2024

Signal Adversarial Examples Generation for Signal Detection Network via White-Box Attack

With the development and application of deep learning in signal detection tasks, the vulnerability of neural networks to adversarial attacks has also become a security threat to signal detection networks. This paper defines a signal adversarial examples generation model for signal detection network from the perspective of adding perturbations to the signal. The model uses the inequality relationship of L2-norm between time domain and time-frequency domain to constrain the energy of signal perturbations. Building upon this model, we propose a method for generating signal adversarial examples utilizing gradient-based attacks and Short-Time Fourier Transform. The experimental results show that under the constraint of signal perturbation energy ratio less than 3%, our adversarial attack resulted in a 28.1% reduction in the mean Average Precision (mAP), a 24.7% reduction in recall, and a 30.4% reduction in precision of the signal detection network. Compared to random noise perturbation of equivalent intensity, our adversarial attack demonstrates a significant attack effect.

cs.CV↗

arXiv2024

Scalable Multi-Objective Optimization for Robust Traffic Signal Control in Uncertain Environments

Intelligent traffic signal control is essential to modern urban management, with important impacts on economic efficiency, environmental sustainability, and quality of daily life. However, in current decades, it continues to pose significant challenges in managing large-scale traffic networks, coordinating intersections, and ensuring robustness under uncertain traffic conditions. This paper presents a scalable multi-objective optimization approach for robust traffic signal control in dynamic and uncertain urban environments. A multi-objective optimization model is proposed in this paper, which incorporates stochastic variables and probabilistic traffic patterns to capture traffic flow dynamics and uncertainty. We propose an algorithm named Adaptive Hybrid Multi-Objective Optimization Algorithm (AHMOA), which addresses the uncertainties of city traffic, including network-wide signal coordination, fluctuating patterns, and environmental impacts. AHMOA simultaneously optimizes multiple objectives, such as average delay, network stability, and system robustness, while adapting to unpredictable changes in traffic. The algorithm combines evolutionary strategies with an adaptive mechanism to balance exploration and exploitation, and incorporates a memory-based evaluation mechanism to leverage historical traffic data. Simulations are conducted in different cities including Manhattan, Paris, Sao Paulo, and Istanbul. The experimental results demonstrate that AHMOA consistently outperforms several state-of-the-art algorithms and the algorithm is competent to provide scalable, robust Pareto optimal solutions for managing complex traffic systems under uncertain environments.

cs.NE↗

arXiv2024

Increased resistance to photooxidation in Dion-Jacobson lead halide perovskites -- implication for perovskite device stability

2D metal halide perovskites have enabled significant stability improvements in perovskite devices, particularly in resistance to moisture. However, some 2D perovskites are even more susceptible to photooxidation compared to 3D perovskites. This is particularly true for more commonly investigated Ruddlesden-Popper (RP) perovskites that exhibit increased susceptibility to photoinduced degradation compared to Dion-Jacobson (DJ) perovskites. Comparisons between different RP and DJ perovskites reveal that this phenomenon cannot be explained by commonly proposed differences in superoxide ion generation, interlayer distance and lattice structural rigidity differences. Instead, the resistance to photooxidation of DJ perovskites can be attributed to decreased likelihood of double deprotonation events (compared to single deprotonation events in RP perovskites) required for the loss of organic cations and the perovskite decomposition. Consequently, DJ perovskites are less susceptible to oxidative degradation (both photo- and electrochemically induced), which leads to improved operational stability of solar cells based on these materials.

cond-mat.mtrl-sci↗

arXiv2024

RealisDance: Equip controllable character animation with realistic hands

Controllable character animation is an emerging task that generates character videos controlled by pose sequences from given character images. Although character consistency has made significant progress via reference UNet, another crucial factor, pose control, has not been well studied by existing methods yet, resulting in several issues: 1) The generation may fail when the input pose sequence is corrupted. 2) The hands generated using the DWPose sequence are blurry and unrealistic. 3) The generated video will be shaky if the pose sequence is not smooth enough. In this paper, we present RealisDance to handle all the above issues. RealisDance adaptively leverages three types of poses, avoiding failed generation caused by corrupted pose sequences. Among these pose types, HaMeR provides accurate 3D and depth information of hands, enabling RealisDance to generate realistic hands even for complex gestures. Besides using temporal attention in the main UNet, RealisDance also inserts temporal attention into the pose guidance network, smoothing the video from the pose condition aspect. Moreover, we introduce pose shuffle augmentation during training to further improve generation robustness and video smoothness. Qualitative experiments demonstrate the superiority of RealisDance over other existing methods, especially in hand quality.

cs.CV↗

arXiv2024

ICGMM: CXL-enabled Memory Expansion with Intelligent Caching Using Gaussian Mixture Model

Compute Express Link (CXL) emerges as a solution for wide gap between computational speed and data communication rates among host and multiple devices. It fosters a unified and coherent memory space between host and CXL storage devices such as such as Solid-state drive (SSD) for memory expansion, with a corresponding DRAM implemented as the device cache. However, this introduces challenges such as substantial cache miss penalties, sub-optimal caching due to data access granularity mismatch between the DRAM "cache" and SSD "memory", and inefficient hardware cache management. To address these issues, we propose a novel solution, named ICGMM, which optimizes caching and eviction directly on hardware, employing a Gaussian Mixture Model (GMM)-based approach. We prototype our solution on an FPGA board, which demonstrates a noteworthy improvement compared to the classic Least Recently Used (LRU) cache strategy. We observe a decrease in the cache miss rate ranging from 0.32% to 6.14%, leading to a substantial 16.23% to 39.14% reduction in the average SSD access latency. Furthermore, when compared to the state-of-the-art Long Short-Term Memory (LSTM)-based cache policies, our GMM algorithm on FPGA showcases an impressive latency reduction of over 10,000 times. Remarkably, this is achieved while demanding much fewer hardware resources.

cs.AR↗

arXiv2024

VideoQA in the Era of LLMs: An Empirical Study

Video Large Language Models (Video-LLMs) are flourishing and has advanced many video-language tasks. As a golden testbed, Video Question Answering (VideoQA) plays pivotal role in Video-LLM developing. This work conducts a timely and comprehensive study of Video-LLMs' behavior in VideoQA, aiming to elucidate their success and failure modes, and provide insights towards more human-like video understanding and question answering. Our analyses demonstrate that Video-LLMs excel in VideoQA; they can correlate contextual cues and generate plausible responses to questions about varied video contents. However, models falter in handling video temporality, both in reasoning about temporal content ordering and grounding QA-relevant temporal moments. Moreover, the models behave unintuitively - they are unresponsive to adversarial video perturbations while being sensitive to simple variations of candidate answers and questions. Also, they do not necessarily generalize better. The findings demonstrate Video-LLMs' QA capability in standard condition yet highlight their severe deficiency in robustness and interpretability, suggesting the urgent need on rationales in Video-LLM developing.

cs.CV↗

arXiv2024

On the Role of Long-tail Knowledge in Retrieval Augmented Large Language Models

Retrieval augmented generation (RAG) exhibits outstanding performance in promoting the knowledge capabilities of large language models (LLMs) with retrieved documents related to user queries. However, RAG only focuses on improving the response quality of LLMs via enhancing queries indiscriminately with retrieved information, paying little attention to what type of knowledge LLMs really need to answer original queries more accurately. In this paper, we suggest that long-tail knowledge is crucial for RAG as LLMs have already remembered common world knowledge during large-scale pre-training. Based on our observation, we propose a simple but effective long-tail knowledge detection method for LLMs. Specifically, the novel Generative Expected Calibration Error (GECE) metric is derived to measure the ``long-tailness'' of knowledge based on both statistics and semantics. Hence, we retrieve relevant documents and infuse them into the model for patching knowledge loopholes only when the input query relates to long-tail knowledge. Experiments show that, compared to existing RAG pipelines, our method achieves over 4x speedup in average inference time and consistent performance improvement in downstream tasks.

cs.IR↗

arXiv2024

UniPSDA: Unsupervised Pseudo Semantic Data Augmentation for Zero-Shot Cross-Lingual Natural Language Understanding

Cross-lingual representation learning transfers knowledge from resource-rich data to resource-scarce ones to improve the semantic understanding abilities of different languages. However, previous works rely on shallow unsupervised data generated by token surface matching, regardless of the global context-aware semantics of the surrounding text tokens. In this paper, we propose an Unsupervised Pseudo Semantic Data Augmentation (UniPSDA) mechanism for cross-lingual natural language understanding to enrich the training data without human interventions. Specifically, to retrieve the tokens with similar meanings for the semantic data augmentation across different languages, we propose a sequential clustering process in 3 stages: within a single language, across multiple languages of a language family, and across languages from multiple language families. Meanwhile, considering the multi-lingual knowledge infusion with context-aware semantics while alleviating computation burden, we directly replace the key constituents of the sentences with the above-learned multi-lingual family knowledge, viewed as pseudo-semantic. The infusion process is further optimized via three de-biasing techniques without introducing any neural parameters. Extensive experiments demonstrate that our model consistently improves the performance on general zero-shot cross-lingual natural language understanding tasks, including sequence classification, information extraction, and question answering.

cs.CL↗

arXiv2024

KEHRL: Learning Knowledge-Enhanced Language Representations with Hierarchical Reinforcement Learning

Knowledge-enhanced pre-trained language models (KEPLMs) leverage relation triples from knowledge graphs (KGs) and integrate these external data sources into language models via self-supervised learning. Previous works treat knowledge enhancement as two independent operations, i.e., knowledge injection and knowledge integration. In this paper, we propose to learn Knowledge-Enhanced language representations with Hierarchical Reinforcement Learning (KEHRL), which jointly addresses the problems of detecting positions for knowledge injection and integrating external knowledge into the model in order to avoid injecting inaccurate or irrelevant knowledge. Specifically, a high-level reinforcement learning (RL) agent utilizes both internal and prior knowledge to iteratively detect essential positions in texts for knowledge injection, which filters out less meaningful entities to avoid diverting the knowledge learning direction. Once the entity positions are selected, a relevant triple filtration module is triggered to perform low-level RL to dynamically refine the triples associated with polysemic entities through binary-valued actions. Experiments validate KEHRL's effectiveness in probing factual knowledge and enhancing the model's performance on various natural language understanding tasks.

cs.CL↗

arXiv2024

DAFNet: Dynamic Auxiliary Fusion for Sequential Model Editing in Large Language Models

Recently, while large language models (LLMs) have demonstrated impressive results, they still suffer from hallucination, i.e., the generation of false information. Model editing is the task of fixing factual mistakes in LLMs; yet, most previous works treat it as a one-time task, paying little attention to ever-emerging mistakes generated by LLMs. We address the task of sequential model editing (SME) that aims to rectify mistakes continuously. A Dynamic Auxiliary Fusion Network (DAFNet) is designed to enhance the semantic interaction among the factual knowledge within the entire sequence, preventing catastrophic forgetting during the editing process of multiple knowledge triples. Specifically, (1) for semantic fusion within a relation triple, we aggregate the intra-editing attention flow into auto-regressive self-attention with token-level granularity in LLMs. We further leverage multi-layer diagonal inter-editing attention flow to update the weighted representations of the entire sequence-level granularity. (2) Considering that auxiliary parameters are required to store the knowledge for sequential editing, we construct a new dataset named \textbf{DAFSet}, fulfilling recent, popular, long-tail and robust properties to enhance the generality of sequential editing. Experiments show DAFNet significantly outperforms strong baselines in single-turn and sequential editing. The usage of DAFSet also consistently improves the performance of other auxiliary network-based methods in various scenarios

cs.CL↗

arXiv2024

Lifelong Knowledge Editing for LLMs with Retrieval-Augmented Continuous Prompt Learning

Model editing aims to correct outdated or erroneous knowledge in large language models (LLMs) without the need for costly retraining. Lifelong model editing is the most challenging task that caters to the continuous editing requirements of LLMs. Prior works primarily focus on single or batch editing; nevertheless, these methods fall short in lifelong editing scenarios due to catastrophic knowledge forgetting and the degradation of model performance. Although retrieval-based methods alleviate these issues, they are impeded by slow and cumbersome processes of integrating the retrieved knowledge into the model. In this work, we introduce RECIPE, a RetriEval-augmented ContInuous Prompt lEarning method, to boost editing efficacy and inference efficiency in lifelong learning. RECIPE first converts knowledge statements into short and informative continuous prompts, prefixed to the LLM's input query embedding, to efficiently refine the response grounded on the knowledge. It further integrates the Knowledge Sentinel (KS) that acts as an intermediary to calculate a dynamic threshold, determining whether the retrieval repository contains relevant knowledge. Our retriever and prompt encoder are jointly trained to achieve editing properties, i.e., reliability, generality, and locality. In our experiments, RECIPE is assessed extensively across multiple LLMs and editing datasets, where it achieves superior editing performance. RECIPE also demonstrates its capability to maintain the overall performance of LLMs alongside showcasing fast editing and inference speed.

cs.CL↗

arXiv2024

General Procedure to Provide High-Probability Guarantees for Stochastic Saddle Point Problems

This paper considers smooth strongly convex and strongly concave (SC-SC) stochastic saddle point (SSP) problems. Suppose there is an arbitrary oracle that in expectation returns an $\epsilon$-solution in the sense of certain gaps, which can be the duality gap or its weaker variants. We propose a general PB-SSP framework to guarantee an $\epsilon$ small duality gap solution with high probability via only $\mathcal{O}\big(\log \frac{1}{p}\cdot\text{poly}(\log \kappa)\big)$ calls of this oracle, where $p\in(0,1)$ is the confidence level and $\kappa$ is the condition number. When applied to the sample average approximation (SAA) oracle, in addition to equipping the solution with high probability, our approach even improves the sample complexity by a factor of $\text{poly}(\kappa)$, since the high-probability argument enables us to circumvent some key difficulties of the uniform stability analysis of SAA.

math.OC↗

arXiv2024

R4: Reinforced Retriever-Reorder-Responder for Retrieval-Augmented Large Language Models

Retrieval-augmented large language models (LLMs) leverage relevant content retrieved by information retrieval systems to generate correct responses, aiming to alleviate the hallucination problem. However, existing retriever-responder methods typically append relevant documents to the prompt of LLMs to perform text generation tasks without considering the interaction of fine-grained structural semantics between the retrieved documents and the LLMs. This issue is particularly important for accurate response generation as LLMs tend to "lose in the middle" when dealing with input prompts augmented with lengthy documents. In this work, we propose a new pipeline named "Reinforced Retriever-Reorder-Responder" (R$^4$) to learn document orderings for retrieval-augmented LLMs, thereby further enhancing their generation abilities while the large numbers of parameters of LLMs remain frozen. The reordering learning process is divided into two steps according to the quality of the generated responses: document order adjustment and document representation enhancement. Specifically, document order adjustment aims to organize retrieved document orderings into beginning, middle, and end positions based on graph attention learning, which maximizes the reinforced reward of response quality. Document representation enhancement further refines the representations of retrieved documents for responses of poor quality via document-level gradient adversarial learning. Extensive experiments demonstrate that our proposed pipeline achieves better factual question-answering performance on knowledge-intensive tasks compared to strong baselines across various public datasets. The source codes and trained models will be released upon paper acceptance.

cs.CL↗