arXiv subjects

Tingwen Huang

Publications and source records attributed to Tingwen Huang.

At least 19 recordsLinked to original sources

arXiv2026

Towards Compact Unified Multimodal Tracking: Synergizing Knowledge Distillation with Structural Pruning

Unified multimodal object tracking has achieved remarkable robustness by leveraging complementary sensor data (e.g., RGB, Thermal, Depth), yet the heavy computational burden of state-of-the-art models hinders their deployment on resource-constrained edge devices. In this work, we identify the prediction head as a critical but often overlooked efficiency bottleneck. By strategically streamlining the decoder architecture, we unlock the potential for real-time inference but simultaneously introduce a capacity gap between the lightweight student and the heavy teacher. To resolve this, we conduct a systematic analysis of 17 distillation strategies and introduce a Dual-Alignment Distillation framework. Our key insight is that effective compression requires decoupling knowledge transfer into two complementary streams: (1) Spatial Representation Alignment, which employs feature distillation to sharpen the student's spatial focus on foreground targets ("Where to track"); and (2) Semantic Distribution Alignment, which utilizes logit-based distillation to align decision boundaries and transfer discriminative dark knowledge ("What to track"). Extensive experiments across five benchmarks demonstrate that our approach significantly outperforms complex state-of-the-art methods. Notably, our distilled model achieves 91.5% MPR on RGBT234 and operates at 54 FPS on a single RTX 4090, representing a 5x speedup over the teacher model while maintaining superior accuracy.

cs.CV↗

arXiv2026

Prox-DBRO-VR: A Unified Analysis on Byzantine-Resilient Decentralized Stochastic Composite Optimization with Variance Reduction and Non-Asymptotic Convergence Rates

Decentralized stochastic gradient algorithms efficiently solve large-scale finite-sum optimization problems when all agents in the network are reliable. However, most of these algorithms are not resilient to adverse conditions, such as malfunctioning agents, software bugs, and cyber attacks. This paper aims to handle a class of general composite optimization problems over multi-agent systems (MASs) in the presence of an unknown number of Byzantine agents. Building on a resilient aggregation mechanism and the proximal-gradient mapping method, a Byzantine-resilient decentralized stochastic proximal-gradient algorithmic framework is proposed, dubbed Prox-DBRO-VR, which achieves an optimization and control goal using only local computations and communications. To asymptotically reduce the noise variance arising from local gradient estimation and accelerate the convergence, we incorporate two localized variance-reduced (VR) techniques (SAGA and LSVRG) into Prox-DBRO-VR to design Prox-DBRO-SAGA and Prox-DBRO-LSVRG. By analyzing the contraction relationships among the gradient-learning error, resilient consensus condition, and convergence error in a unified theoretical framework, it is proved that both Prox-DBRO-SAGA and Prox-DBRO-LSVRG, with a well-designed constant (resp., decaying) step-size, converge linearly (resp., sub-linearly) inside an error ball around the optimal solution to the original problem under standard assumptions. A trade-off between convergence accuracy and Byzantine resilience in both linear and sub-linear cases is also characterized. In numerical experiments, the effectiveness and practicability of the proposed algorithms are manifested via resolving a decentralized sparse machine-learning problem under various Byzantine attacks.

math.OC↗

arXiv2026

RED-SEGA:Resilient Decentralized Stochastic Proximal Optimization with Gradient Sketching over Time-Varying Networks

Variance reduction is indispensable in Byzantine-resilient decentralized stochastic optimization over multi-agent systems (MASs) for its ability to mitigate gradient noise and thereby enhance the resilient aggregation process. However, most existing Byzantine-resilient decentralized variance-reduced (VR) stochastic gradient algorithms rely on random data sampling, which proves inefficient in data-scarce yet high-dimensional tasks, for instance, image deblurring. This paper pursues an alternative technical line that achieves variance reduction via gradient sketching. To this end, we first formulate a class of structural risk minimization (SRM) problems, where the local objectives are not necessarily decomposable and their gradients may be unavailable. To solve the SRM problems in a decentralized manner, we integrate a gradient-sketching technique into decentralized stochastic proximal gradient descent with gossip communication to propose a decentralized VR stochastic gradient algorithm, dubbed Gossip-SEGA.Since Gossip-SEGA does not provide any resilience against Byzantine attacks, a resilient extension of Gossip-SEGA,namely RED-SEGA,is developed via replacing the weighted average in Gossip-SEGA by a norm-penalized approximation. Theoretically, we derive sufficient conditions for both consensus (among reliable agents) and linear convergence rate of RED-SEGA over time-varying networks. The effectiveness and resilience of the proposed algorithms are validated through numerical experiments.

math.OC↗

arXiv2026

Low-Rank Tensor Completion Based on Fractional Regularization with Ky Fan p-k Norm

This paper addresses low-rank tensor completion (LRTC) by proposing a novel nonconvex surrogate, namely the ratio of the tensor nuclear norm to the tensor Ky Fan p-k norm (TNPK), to accurately approximate the tensor tubal rank. The TNPK possesses appealing properties, including scale invariance, parameter flexibility, and the existence of closed-form solutions under specific choices of p and k. With specific parameter settings of p and k, it reduces to the ratio of the tensor nuclear norm to the tensor Ky Fan k norm (TNK) or the ratio of the tensor nuclear norm to the tensor Frobenius norm (TNF). We construct a LRTC model and, under the tensor null space property (NSP), prove that low-rank tensors are local minimizers of the proposed model. Moreover, we derive the proximal operator of the Ky Fan p-k inverse-norm and further develop an efficient alternating direction method of multipliers (ADMM) algorithm with guaranteed subsequential convergence under mild conditions. Extensive experiments on synthetic and real-world datasets validate the superior performance of our method against state-of-the-art competitors.

cs.CV↗

arXiv2026

Federated Knowledge Distillation for Multi-Model Architectures Lithography Hotspot Detection

As a special type of multimedia data, Lithography Hotspot Detection (LHD) training often requires stronger privacy protection than conventional multimedia data, and federated learning provides a promising potential solution to this challenge. However, existing approaches rely solely on either parameter aggregation or Knowledge Distillation (KD), failing to fully exploit the potential of collaborative learning. To address this, we propose FedKD-hybrid, a novel framework that synergizes the strengths of both paradigms. Specifically, FedKD-hybrid utilizes a public dataset to facilitate consensus, where clients exchange both parameters of agreed-upon layers and logits. This hybrid information is aggregated to refine local models, enhancing knowledge transfer. Extensive experiments on ICCAD-2012 and real-world FAB datasets demonstrate that FedKD-hybrid consistently outperforms state-of-the-art methods in both effectiveness and robustness.

cs.LG↗

arXiv2026

Efficient state transition algorithm with guaranteed optimality

The state transition algorithm (STA), as an intelligent optimization method grounded in constructivist learning, has been demonstrated to be highly effective in solving complex optimization problems. However, the standard STA suffers from slow convergence, particularly in the later stages when dealing with flat landscapes. Additionally, users are required to set the maximum number of iterations based on intuition. To address these issues, an enhanced STA with guaranteed optimality is introduced. This improvement involves three key components. First, novel translation transformations, inspired by predictive modeling, are developed to generate a broader set of candidate solutions by leveraging historical data. Second, adaptive parameter control strategies are incorporated to accelerate convergence. Finally, a dedicated termination condition is designed to ensure that the algorithm converges at the optimal solution, analogous to the zero gradient condition in mathematical programming. The comprehensive experimental results validate the effectiveness and superiority of the proposed method. The source codes for ESTA and EXSTA will be publicly available at: https://github.com/tiezhongyu2005/ESTA.

math.OC↗

arXiv2026

V-STC: A Time-Efficient Multi-Vehicle Coordinated Trajectory Planning Approach

Coordinating the motions of multiple autonomous vehicles (AVs) requires planning frameworks that ensure safety while making efficient use of space and time. This paper presents a new approach, termed variable-time-step spatio-temporal corridor (V-STC), that enhances the temporal efficiency of multi-vehicle coordination. An optimization model is formulated to construct a V-STC for each AV, in which both the spatial configuration of the corridor cubes and their time durations are treated as decision variables. By allowing the corridor's spatial position and time step to vary, the constructed V-STC reduces the overall temporal occupancy of each AV while maintaining collision-free separation in the spatio-temporal domain. Based on the generated V-STC, a dynamically feasible trajectory is then planned independently for each AV. Simulation studies demonstrate that the proposed method achieves safe multi-vehicle coordination and yields more time-efficient motion compared with existing STC approaches.

cs.RO↗

arXiv2026

Optimal Decentralized Dynamic Energy Management over Asynchronous Peer-to-Peer Transactive Networks via Operator Splitting

Peer-to-peer (P2P) energy management facilitates decentralized resource allocation among prosumers, improving local hosting capacity for renewables and minimizing energy expenditures while ensuring data privacy through distributed coordination. However, conventional P2P energy management methods are confined to synchronous scheduling paradigms, creating synchronization bottlenecks that fundamentally conflict with the dynamic and decentralized nature of P2P energy management tasks. To bridge this gap, this paper focuses on resolving a class of dynamic energy management problems over asynchronous P2P (Asyn-P2P) transactive networks. We first recast the dynamic energy management problems into a saddle-point problem, and then propose a synchronous decentralized dynamic energy management algorithm, dubbed Syn-DYNA,based on operator splitting theory. To eliminate the global synchronization clock in Syn-DYNA, we introduce a random activation scheme, together with local buffers for latest state tracking, to develop an asynchronous variant of Syn-DYNA, namely Asyn-DYNA. Based on monotone operator theory, theoretical analysis proves a non-asymptotic linear convergence rate for Syn-DYNA and establishes the almost sure convergence ofAsyn-DYNA. Numerical experiments validate effectiveness of Syn-DYNA and Asyn-DYNA algorithms by tackling a dynamic energy management task over P2P transactive networks.

eess.SY↗

arXiv2026

SpikeSMOKE: Spiking Neural Networks for Monocular 3D Object Detection with Cross-Scale Gated Coding

With the wide application of 3D object detection in some fields such as autonomous driving, its energy consumption is constantly increasing, making the research on low-power consumption alternatives a key research area. The spiking neural networks (SNNs), possessing low-power consumption characteristics, offer a novel solution for this research. Consequently, we apply SNNs to monocular 3D object detection and propose the SpikeSMOKE architecture, which represents a new attempt at low-power monocular 3D object detection. It's well known that the discrete signals of SNNs can lead to information loss compared to artificial neural networks (ANNs), which limits their feature representation capabilities. To solve this problem, inspired by the synaptic filtering mechanism of biological neurons, we propose a new Cross-Scale Gating Coding Mechanism (CSGC), which can enhance feature representation by combining cross-scale fusion of attentional methods and gated filtering mechanisms. In addition, to reduce the computation and accelerate training, we present a novel light-weight residual block that can maintain spiking computing paradigm and the highest possible detection performance. Our method is effective on the KITTI, NuScenes-mini and CIFAR10/100 datasets. Compared to the baseline SpikeSMOKE under the 3D Object Detection, the proposed SpikeSMOKE with CSGC can achieve 11.78 (+2.82, Easy), 10.69 (+3.2, Moderate), and 10.48 (+3.17, Hard) on the KITTI autonomous driving dataset by AP|R11 at 0.7 IoU threshold, respectively. It is worth noting that the results of SpikeSMOKE can significantly reduce energy consumption compared with the results of SMOKE. And SpikeSMOKE-L (lightweight) can further reduce the amount of parameters by 3 times and computation by 10 times compared to SMOKE.

cs.CV↗

arXiv2026

Hit-RAG: Learning to Reason with Long Contexts via Preference Alignment

Despite the promise of Retrieval-Augmented Generation in grounding Multimodal Large Language Models with external knowledge, the transition to extensive contexts often leads to significant attention dilution and reasoning hallucinations. The surge in information density causes critical evidence to be submerged by voluminous noise, which complicates the discernment of relevant fragments within a dense input. In this paper, we propose \textbf{Hit-RAG}, a multi-stage preference alignment framework designed to resolve these cognitive bottlenecks through a progressive optimization pipeline. Our approach systematically refines the utilization of external evidence via three distinct stages. First, Supervised Fine-tuning establishes baseline context awareness to minimize information neglect. Next, Discriminative Preference Alignment enhances robustness against misleading distractors. Finally, Group-Relative Policy Optimization stabilizes logical synthesis to prevent reasoning collapse. Extensive evaluations on eight benchmarks demonstrate that Hit-RAG consistently yields substantial performance gains, enabling models to bridge the gap between context acquisition and accurate reasoning while surpassing much larger counterparts in long-context scenarios.

cs.CL↗

arXiv2026

MO-MIX: Multi-Objective Multi-Agent Cooperative Decision-Making With Deep Reinforcement Learning

Deep reinforcement learning (RL) has been applied extensively to solve complex decision-making problems. In many real-world scenarios, tasks often have several conflicting objectives and may require multiple agents to cooperate, which are the multi-objective multi-agent decision-making problems. However, only few works have been conducted on this intersection. Existing approaches are limited to separate fields and can only handle multi-agent decision-making with a single objective, or multi-objective decision-making with a single agent. In this paper, we propose MO-MIX to solve the multi-objective multi-agent reinforcement learning (MOMARL) problem. Our approach is based on the centralized training with decentralized execution (CTDE) framework. A weight vector representing preference over the objectives is fed into the decentralized agent network as a condition for local action-value function estimation, while a mixing network with parallel architecture is used to estimate the joint action-value function. In addition, an exploration guide approach is applied to improve the uniformity of the final non-dominated solutions. Experiments demonstrate that the proposed method can effectively solve the multi-objective multi-agent cooperative decision-making problem and generate an approximation of the Pareto set. Our approach not only significantly outperforms the baseline method in all four kinds of evaluation metrics, but also requires less computational cost.

cs.AI↗

arXiv2026

Multi-Dimensional Visual Data Recovery: Scale-Aware Tensor Modeling and Accelerated Randomized Computation

The recently proposed fully-connected tensor network (FCTN) decomposition has demonstrated significant advantages in correlation characterization and transpositional invariance, and has achieved notable achievements in multi-dimensional data processing and analysis. However, existing multi-dimensional data recovery methods leveraging FCTN decomposition still have room for further enhancement, particularly in computational efficiency and modeling capability. To address these issues, we first propose a FCTN-based generalized nonconvex regularization paradigm from the perspective of gradient mapping. Then, reliable and scalable multi-dimensional data recovery models are investigated, where the model formulation is shifted from unquantized observations to coarse-grained quantized observations. Based on the alternating direction method of multipliers (ADMM) framework, we derive efficient optimization algorithms with convergence guarantees to solve the formulated models. To alleviate the computational bottleneck encountered when processing large-scale multi-dimensional data, fast and efficient randomized compression algorithms are devised in virtue of sketching techniques in numerical linear algebra. These dimensionality-reduction techniques serve as the computational acceleration core of our proposed algorithm framework. Theoretical results on approximation error upper bounds and convergence analysis for the proposed method are derived. Extensive numerical experiments illustrate the effectiveness and superiority of the proposed algorithm over other state-of-the-art methods in terms of quantitative metrics, visual quality, and running time.

cs.LG↗

arXiv2026

Non-Trivial Consensus on Directed Matrix-Weighted Networks with Cooperative and Antagonistic Interactions

This paper investigates the non-trivial consensus problem on directed signed matrix-weighted networks\textemdash a novel convergence state that has remained largely unexplored despite prior studies on bipartite consensus and trivial consensus. Notably, we first prove that for directed signed matrix-weighted networks, every eigenvalue of the grounded Laplacians has positive real part under certain conditions. This key finding ensures the global asymptotic convergence of systems states to the null spaces of signed matrix-weighted Laplacians, providing a foundational tool for analyzing dynamics on rooted signed matrix-weighted networks. To achieve non-trivial consensus, we propose a systematic approach involving the strategic selection of informed agents, careful design of external signals, and precise determination of coupling terms. Crucially, we derive the lower bounds of the coupling coefficients. Our consensus algorithm operates under milder connectivity conditions, and does not impose restrictions on whether the network is structurally balanced or unbalanced. Moreover, the non-trivial consensus state can be preset arbitrarily as needed. We also carry out the above analysis for undirected networks, with more relaxed conditions on the coupling coefficients comparing to the directed case. This paper further studies non-trivial consensus with switching topologies, and propose the necessary condition for the convergence of switching networks. The work in this paper demonstrates that groups with both cooperative and antagonistic multi-dimensional interactions can achieve consensus, which was previously deemed exclusive to fully cooperative groups.

eess.SY↗

arXiv2025

MMT-ARD: Multimodal Multi-Teacher Adversarial Distillation for Robust Vision-Language Models

Vision-Language Models (VLMs) are increasingly deployed in safety-critical applications, making their adversarial robustness a crucial concern. While adversarial knowledge distillation has shown promise in transferring robustness from teacher to student models, traditional single-teacher approaches suffer from limited knowledge diversity, slow convergence, and difficulty in balancing robustness and accuracy. To address these challenges, we propose MMT-ARD: a Multimodal Multi-Teacher Adversarial Robust Distillation framework. Our key innovation is a dual-teacher knowledge fusion architecture that collaboratively optimizes clean feature preservation and robust feature enhancement. To better handle challenging adversarial examples, we introduce a dynamic weight allocation strategy based on teacher confidence, enabling adaptive focus on harder samples. Moreover, to mitigate bias among teachers, we design an adaptive sigmoid-based weighting function that balances the strength of knowledge transfer across modalities. Extensive experiments on ImageNet and zero-shot benchmarks demonstrate that MMT-ARD improves robust accuracy by +4.32% and zero-shot accuracy by +3.5% on the ViT-B-32 model, while achieving a 2.3x increase in training efficiency over traditional single-teacher methods. These results highlight the effectiveness and scalability of MMT-ARD in enhancing the adversarial robustness of multimodal large models. Our codes are available at https://github.com/itsnotacie/MMT-ARD.

cs.CV↗

arXiv2025

SGLP: A Similarity Guided Fast Layer Partition Pruning for Compressing Large Deep Models

Layer pruning has emerged as a potent approach to remove redundant layers in the pre-trained network on the purpose of reducing network size and improve computational efficiency. However, existing layer pruning methods mostly overlook the intrinsic connections and inter-dependencies between different layers within complicated deep neural networks. This oversight can result in pruned models that do not preserve the essential characteristics of the pre-trained network as effectively as desired. To address these limitations, we propose a Similarity-Guided Layer Partition (SGLP) Pruning, a novel pruning framework that exploits representation similarity to guide efficient and informed layer removal for compressing large deep models. Our method begins by employing Centered Kernel Alignment (CKA) to quantify representational similarity between layers, uncovering structural patterns within the network. We then apply Fisher Optimal Segmentation on the similarity matrix to partition the network into semantically coherent layer segments. This segmentation allows pruning decisions to respect layer interdependencies and preserve essential knowledge. Within each segment, we introduce a fine-tuning-free importance evaluation using GradNorm, identifying and removing redundant layers in a targeted, segment-wise manner. Experimental results on both image classification tasks and large language models (LLMs) demonstrate that our proposed SGLP outperforms the state-of-the-art methods in accuracy and efficiency. Our approach achieves significant model compression with minimal performance degradation, making it well-suited for deployment in resource-limited environments.

cs.LG↗

arXiv2025

Nash equilibrium seeking in coalition games for multiple Euler-Lagrange systems: Analysis and application to USV swarm confrontation

This paper addresses a class of Nash equilibrium (NE) seeking problems in coalition games involving both local and coupling constraints for multiple Euler-Lagrange (EL) systems subject to disturbances of unknown bounds. Within each coalition, agents cooperatively minimize a shared cost function while competing against other coalitions. A distributed strategy is proposed to seek the NE under informational constraints, where each agent has access only to its own action, cost function, and constraint parameters. In the proposed distributed NE seeking strategy, adaptive techniques are combined with sign functions to handle model uncertainties and disturbances with unknown bounds in the EL systems. To deal with the Lagrange multipliers associated with local and coupling constraints, primal-dual techniques are integrated with consensus protocols. Additionally, a dynamic average consensus algorithm is employed to estimate the gradient of the coalition cost function, while a leader-following protocol is utilized to estimate the actions of other agents. Under standard convexity and graph-connectivity assumptions, global convergence of the closed-loop EL system to the NE is established. As an illustrative application, a swarm confrontation of unmanned surface vehicles involving formation, encirclement, and interception tasks is modeled within the coalition game framework, and numerical simulations are conducted under this model to validate the theoretical results.

eess.SY↗

arXiv2025

MPQ-DMv2: Flexible Residual Mixed Precision Quantization for Low-Bit Diffusion Models with Temporal Distillation

Diffusion models have demonstrated remarkable performance on vision generation tasks. However, the high computational complexity hinders its wide application on edge devices. Quantization has emerged as a promising technique for inference acceleration and memory reduction. However, existing quantization methods do not generalize well under extremely low-bit (2-4 bit) quantization. Directly applying these methods will cause severe performance degradation. We identify that the existing quantization framework suffers from the outlier-unfriendly quantizer design, suboptimal initialization, and optimization strategy. We present MPQ-DMv2, an improved \textbf{M}ixed \textbf{P}recision \textbf{Q}uantization framework for extremely low-bit \textbf{D}iffusion \textbf{M}odels. For the quantization perspective, the imbalanced distribution caused by salient outliers is quantization-unfriendly for uniform quantizer. We propose \textit{Flexible Z-Order Residual Mixed Quantization} that utilizes an efficient binary residual branch for flexible quant steps to handle salient error. For the optimization framework, we theoretically analyzed the convergence and optimality of the LoRA module and propose \textit{Object-Oriented Low-Rank Initialization} to use prior quantization error for informative initialization. We then propose \textit{Memory-based Temporal Relation Distillation} to construct an online time-aware pixel queue for long-term denoising temporal information distillation, which ensures the overall temporal consistency between quantized and full-precision model. Comprehensive experiments on various generation tasks show that our MPQ-DMv2 surpasses current SOTA methods by a great margin on different architectures, especially under extremely low-bit widths.

cs.CV↗

arXiv2025

IGDNet: Zero-Shot Robust Underexposed Image Enhancement via Illumination-Guided and Denoising

Current methods for restoring underexposed images typically rely on supervised learning with paired underexposed and well-illuminated images. However, collecting such datasets is often impractical in real-world scenarios. Moreover, these methods can lead to over-enhancement, distorting well-illuminated regions. To address these issues, we propose IGDNet, a Zero-Shot enhancement method that operates solely on a single test image, without requiring guiding priors or training data. IGDNet exhibits strong generalization ability and effectively suppresses noise while restoring illumination. The framework comprises a decomposition module and a denoising module. The former separates the image into illumination and reflection components via a dense connection network, while the latter enhances non-uniformly illuminated regions using an illumination-guided pixel adaptive correction method. A noise pair is generated through downsampling and refined iteratively to produce the final result. Extensive experiments on four public datasets demonstrate that IGDNet significantly improves visual quality under complex lighting conditions. Quantitative results on metrics like PSNR (20.41dB) and SSIM (0.860dB) show that it outperforms 14 state-of-the-art unsupervised methods. The code will be released soon.

cs.CV↗