Searcharxiv⌕ Search

arXiv subjects

Ming Tang

Publications and source records attributed to Ming Tang.

At least 127 records · Page 7Linked to original sources
arXiv2024

BFRFormer: Transformer-based generator for Real-World Blind Face Restoration

Blind face restoration is a challenging task due to the unknown and complex degradation. Although face prior-based methods and reference-based methods have recently demonstrated high-quality results, the restored images tend to contain over-smoothed results and lose identity-preserved details when the degradation is severe. It is observed that this is attributed to short-range dependencies, the intrinsic limitation of convolutional neural networks. To model long-range dependencies, we propose a Transformer-based blind face restoration method, named BFRFormer, to reconstruct images with more identity-preserved details in an end-to-end manner. In BFRFormer, to remove blocking artifacts, the wavelet discriminator and aggregated attention module are developed, and spectral normalization and balanced consistency regulation are adaptively applied to address the training instability and over-fitting problem, respectively. Extensive experiments show that our method outperforms state-of-the-art methods on a synthetic dataset and four real-world datasets. The source code, Casia-Test dataset, and pre-trained models are released at https://github.com/s8Znk/BFRFormer.

cs.CV↗
arXiv2024

Multimode fiber speckle Stokes polarimeter

The detection of the state of polarization (SOP) of light is essential for many optical applications. However, it is a challenge for cost-effective SOP measurement due to the complexity of conventional methods and poor transferability of new methods. Here, we propose a straightforward, low-cost and portable SOP measurement system based on the multimode fiber speckle. Convolutional neural network is utilized to establish the mapping relationship between speckle and Stokes parameters. The lowest root mean square error of the estimated SOP on Poincare sphere can be 0.0042. This method is distinguished by its low cost, clear structure and applicability to different wavelengths with high precision. The proposed method is of great value in polarization-related applications.

physics.optics↗
arXiv2024

Accelerate Microstructure Evolution Simulation Using Graph Neural Networks with Adaptive Spatiotemporal Resolution

Surrogate models driven by sizeable datasets and scientific machine-learning methods have emerged as an attractive microstructure simulation tool with the potential to deliver predictive microstructure evolution dynamics with huge savings in computational costs. Taking 2D and 3D grain growth simulations as an example, we present a completely overhauled computational framework based on graph neural networks with not only excellent agreement to both the ground truth phase-field methods and theoretical predictions, but enhanced accuracy and efficiency compared to previous works based on convolutional neural networks. These improvements can be attributed to the graph representation, both improved predictive power and a more flexible data structure amenable to adaptive mesh refinement. As the simulated microstructures coarsen, our method can adaptively adopt remeshed grids and larger timesteps to achieve further speedup. The data-to-model pipeline with training procedures together with the source codes are provided.

cond-mat.mtrl-sci↗
arXiv2023

AnomalyGPT: Detecting Industrial Anomalies Using Large Vision-Language Models

Large Vision-Language Models (LVLMs) such as MiniGPT-4 and LLaVA have demonstrated the capability of understanding images and achieved remarkable performance in various visual tasks. Despite their strong abilities in recognizing common objects due to extensive training datasets, they lack specific domain knowledge and have a weaker understanding of localized details within objects, which hinders their effectiveness in the Industrial Anomaly Detection (IAD) task. On the other hand, most existing IAD methods only provide anomaly scores and necessitate the manual setting of thresholds to distinguish between normal and abnormal samples, which restricts their practical implementation. In this paper, we explore the utilization of LVLM to address the IAD problem and propose AnomalyGPT, a novel IAD approach based on LVLM. We generate training data by simulating anomalous images and producing corresponding textual descriptions for each image. We also employ an image decoder to provide fine-grained semantic and design a prompt learner to fine-tune the LVLM using prompt embeddings. Our AnomalyGPT eliminates the need for manual threshold adjustments, thus directly assesses the presence and locations of anomalies. Additionally, AnomalyGPT supports multi-turn dialogues and exhibits impressive few-shot in-context learning capabilities. With only one normal shot, AnomalyGPT achieves the state-of-the-art performance with an accuracy of 86.1%, an image-level AUC of 94.1%, and a pixel-level AUC of 95.3% on the MVTec-AD dataset. Code is available at https://github.com/CASIA-IVA-Lab/AnomalyGPT.

cs.CV↗
arXiv2023

Fractional Deep Reinforcement Learning for Age-Minimal Mobile Edge Computing

Mobile edge computing (MEC) is a promising paradigm for real-time applications with intensive computational needs (e.g., autonomous driving), as it can reduce the processing delay. In this work, we focus on the timeliness of computational-intensive updates, measured by Age-ofInformation (AoI), and study how to jointly optimize the task updating and offloading policies for AoI with fractional form. Specifically, we consider edge load dynamics and formulate a task scheduling problem to minimize the expected time-average AoI. The uncertain edge load dynamics, the nature of the fractional objective, and hybrid continuous-discrete action space (due to the joint optimization) make this problem challenging and existing approaches not directly applicable. To this end, we propose a fractional reinforcement learning(RL) framework and prove its convergence. We further design a model-free fractional deep RL (DRL) algorithm, where each device makes scheduling decisions with the hybrid action space without knowing the system dynamics and decisions of other devices. Experimental results show that our proposed algorithms reduce the average AoI by up to 57.6% compared with several non-fractional benchmarks.

cs.LG↗
arXiv2023

Fluctuation-based Adaptive Structured Pruning for Large Language Models

Network Pruning is a promising way to address the huge computing resource demands of the deployment and inference of Large Language Models (LLMs). Retraining-free is important for LLMs' pruning methods. However, almost all of the existing retraining-free pruning approaches for LLMs focus on unstructured pruning, which requires specific hardware support for acceleration. In this paper, we propose a novel retraining-free structured pruning framework for LLMs, named FLAP (FLuctuation-based Adaptive Structured Pruning). It is hardware-friendly by effectively reducing storage and enhancing inference speed. For effective structured pruning of LLMs, we highlight three critical elements that demand the utmost attention: formulating structured importance metrics, adaptively searching the global compressed model, and implementing compensation mechanisms to mitigate performance loss. First, FLAP determines whether the output feature map is easily recoverable when a column of weight is removed, based on the fluctuation pruning metric. Then it standardizes the importance scores to adaptively determine the global compressed model structure. At last, FLAP adds additional bias terms to recover the output feature maps using the baseline values. We thoroughly evaluate our approach on a variety of language benchmarks. Without any retraining, our method significantly outperforms the state-of-the-art methods, including LLM-Pruner and the extension of Wanda in structured pruning. The code is released at https://github.com/CASIA-IVA-Lab/FLAP.

cs.CL↗
arXiv2023

Prompt: Probability-Conserved Cross Section Biasing Monte Carlo Particle Transport System

An open source software package for simulating thermal neutron propagation in geometry is presented. In this system, neutron propagation can be treated by either the particle transport method or the ray-tracing method. Supported by an accurate backend scattering physics engine, this system is capable of reproducing neutron scattering experiments in complex geometries and is expected to be used in the areas of instrument characterisation, optimisation and data analysis. In this paper, the relevant theories are briefly introduced. The simulation flow and the user input syntax to control it are provided in detail. Five benchmarking simulations, focusing on different aspects of simulation and scattering techniques, are given to demonstrate the applications of this simulation system. They include an idealised total scattering instrument, a monochromatic powder diffractometer, a neutron guide, a chopper and an imaging setup for complex geometries. Simulated results are benchmarked against experimental data or well-established software packages when appropriate. Good agreements are observed.

physics.comp-ph↗
arXiv2023

Continual Instruction Tuning for Large Multimodal Models

Instruction tuning is now a widely adopted approach to aligning large multimodal models (LMMs) to follow human intent. It unifies the data format of vision-language tasks, enabling multi-task joint training. However, vision-language tasks are constantly being created in practice. Instead of always re-training LMMs when new tasks arrive, continual learning offers flexibility for models to continually and efficiently exploit the evolving data. This work aims to explore the following two questions: 1) Do LMMs still suffer from catastrophic forgetting in continual instruction tuning? 2) Are the existing three classes of continual learning methods still applicable to the continual instruction tuning of LMMs? An extensive study is conducted to address the above questions. First, we establish the first benchmark in this setting and reveal that catastrophic forgetting is still observed when continually instruction-tuning LMMs. However, the multi-task joint instruction tuning can facilitate the model's continual learning ability and mitigate forgetting. Second, we integrate and adapt classic continual learning methods to our context, demonstrating the efficacy of data replay and model expansion strategies across diverse scenarios. In contrast, regularization-based methods only perform well on models that have been jointly instruction-tuned on multiple tasks. Third, we delve into the correlation and forgetting dynamics between vision-language task pairs and propose task-similarity-informed regularization and model expansion methods for continual instruction tuning of LMMs. Experimental results show that our approach consistently boosts the model's performance.

cs.LG↗
arXiv2023

Mitigating Hallucination in Visual Language Models with Visual Supervision

Large vision-language models (LVLMs) suffer from hallucination a lot, generating responses that apparently contradict to the image content occasionally. The key problem lies in its weak ability to comprehend detailed content in a multi-modal context, which can be mainly attributed to two factors in training data and loss function. The vision instruction dataset primarily focuses on global description, and the auto-regressive loss function favors text modeling rather than image understanding. In this paper, we bring more detailed vision annotations and more discriminative vision models to facilitate the training of LVLMs, so that they can generate more precise responses without encounter hallucination. On one hand, we generate image-text pairs with detailed relationship annotations in panoptic scene graph dataset (PSG). These conversations pay more attention on detailed facts in the image, encouraging the model to answer questions based on multi-modal contexts. On the other hand, we integrate SAM and mask prediction loss as auxiliary supervision, forcing the LVLMs to have the capacity to identify context-related objects, so that they can generate more accurate responses, mitigating hallucination. Moreover, to provide a deeper evaluation on the hallucination in LVLMs, we propose a new benchmark, RAH-Bench. It divides vision hallucination into three different types that contradicts the image with wrong categories, attributes or relations, and introduces False Positive Rate as detailed sub-metric for each type. In this benchmark, our approach demonstrates an +8.4% enhancement compared to original LLaVA and achieves widespread performance improvements across other models.

cs.CV↗
arXiv2023

Contagion dynamics in time-varying metapopulation networks with node's activity and attractiveness

The metapopulation network model is effectively used to study the spatial spread of epidemics with individuals mobility. Considering the time-varying nature of individual activity and the preferences for attractive destinations in population mobility, this paper develops a time-varying network model in which activity of a population is correlated with its attractiveness. Based on the model, the spreading processes of the SIR disease on different correlated networks are studied, and global migration thresholds are derived. It is observed that increasing the correlation between activity and attractiveness results in a reduced outbreak threshold but suppresses the disease outbreak size and introduces greater heterogeneity in the spatial distribution of infected individuals. We also investigate the impact of non-pharmacological interventions (self-isolation and self-protection) on the spread of epidemics in different correlation networks. The results show that the simultaneous implementation of these measures is more effective in negatively correlated networks than in positively correlated or non-correlated networks, and the prevalence is reduced significantly. In addition, both self-isolation and self-protection strategies increase the migration threshold of the spreading and thus slow the spread of the epidemic. However, the effectiveness of each strategy in reducing the density of infected populations varies depending on different correlated networks. Self-protection is more effective in positively correlated networks, whereas self-isolation is more effective in negatively correlated networks. These findings contribute to a better understanding of epidemic spreading in large-scale time-varying metapopulation networks and provide insights for epidemic prevention and control.

physics.soc-ph↗
arXiv2023

Pruning-aware Sparse Regularization for Network Pruning

Structural neural network pruning aims to remove the redundant channels in the deep convolutional neural networks (CNNs) by pruning the filters of less importance to the final output accuracy. To reduce the degradation of performance after pruning, many methods utilize the loss with sparse regularization to produce structured sparsity. In this paper, we analyze these sparsity-training-based methods and find that the regularization of unpruned channels is unnecessary. Moreover, it restricts the network's capacity, which leads to under-fitting. To solve this problem, we propose a novel pruning method, named MaskSparsity, with pruning-aware sparse regularization. MaskSparsity imposes the fine-grained sparse regularization on the specific filters selected by a pruning mask, rather than all the filters of the model. Before the fine-grained sparse regularization of MaskSparity, we can use many methods to get the pruning mask, such as running the global sparse regularization. MaskSparsity achieves 63.03%-FLOPs reduction on ResNet-110 by removing 60.34% of the parameters, with no top-1 accuracy loss on CIFAR-10. On ILSVRC-2012, MaskSparsity reduces more than 51.07% FLOPs on ResNet-50, with only a loss of 0.76% in the top-1 accuracy. The code is released at https://github.com/CASIA-IVA-Lab/MaskSparsity. Moreover, we have integrated the code of MaskSparity into a PyTorch pruning toolkit, EasyPruner, at https://gitee.com/casia_iva_engineer/easypruner.

cs.CV↗
arXiv2023

Price of Stability in Quality-Aware Federated Learning

Federated Learning (FL) is a distributed machine learning scheme that enables clients to train a shared global model without exchanging local data. The presence of label noise can severely degrade the FL performance, and some existing studies have focused on algorithm design for label denoising. However, they ignored the important issue that clients may not apply costly label denoising strategies due to them being self-interested and having heterogeneous valuations on the FL performance. To fill this gap, we model the clients' interactions as a novel label denoising game and characterize its equilibrium. We also analyze the price of stability, which quantifies the difference in the system performance (e.g., global model accuracy, social welfare) between the equilibrium outcome and the socially optimal solution. We prove that the equilibrium outcome always leads to a lower global model accuracy than the socially optimal solution does. We further design an efficient algorithm to compute the socially optimal solution. Numerical experiments on MNIST dataset show that the price of stability increases as the clients' data become noisier, calling for an effective incentive mechanism.

cs.LG↗
arXiv2023

Physics-informed neural network for friction-involved nonsmooth dynamics problems

Friction-induced vibration (FIV) is very common in engineering areas. Analysing the dynamic behaviour of systems containing a multiple-contact point frictional interface is an important topic. However, accurately simulating nonsmooth/discontinuous dynamic behaviour due to friction is challenging. This paper presents a new physics-informed neural network approach for solving nonsmooth friction-induced vibration or friction-involved vibration problems. Compared with schemes of the conventional time-stepping methodology, in this new computational framework, the theoretical formulations of nonsmooth multibody dynamics are transformed and embedded in the training process of the neural network. Major findings include that the new framework not only can perform accurate simulation of nonsmooth dynamic behaviour, but also eliminate the need for extremely small time steps typically associated with the conventional time-stepping methodology for multibody systems, thus saving much computation work while maintaining high accuracy. Specifically, four kinds of high-accuracy PINN-based methods are proposed: (1) single PINN; (2) dual PINN; (3) advanced single PINN; (4) advanced dual PINN. Two typical dynamics problems with nonsmooth contact are tested: one is a 1-dimensional contact problem with stick-slip, and the other is a 2-dimensional contact problem considering separation-reattachment and stick-slip oscillation. Both single and dual PINN methods show their advantages in dealing with the 1-dimensional stick-slip problem, which outperforms conventional methods across friction models that are difficult to simulate by the conventional time-stepping method. For the 2-dimensional problem, the capability of the advanced single and advanced dual PINN on accuracy improvement is shown, and they provide good results even in the cases when conventional methods fail.

cs.CE↗
arXiv2023

Transmission of optical communication signals through ring core fiber using perfect vortex beams

Orbital angular momentum can be used to implement high capacity data transmission systems that can be applied for classical and quantum communications. Here we experimentally study the generation and transmission properties of the so-called perfect vortex beams and the Laguerre-Gaussian beams in ring-core optical fibers. Our results show that when using a single preparation stage, the perfect vortex beams present less ring-radius variation that allows coupling of higher optical power into a ring core fiber. These results lead to lower power requirements to establish fiber-based communications links using orbital angular momentum and set the stage for future implementations of high-dimensional quantum communication over space division multiplexing fibers.

physics.optics↗
arXiv2023

FrFT based estimation of linear and nonlinear impairments using Vision Transformer

To comprehensively assess optical fiber communication system conditions, it is essential to implement joint estimation of the following four critical impairments: nonlinear signal-to-noise ratio (SNRNL), optical signal-to-noise ratio (OSNR), chromatic dispersion (CD) and differential group delay (DGD). However, current studies only achieve identifying a limited number of impairments within a narrow range, due to limitations in network capabilities and lack of unified representation of impairments. To address these challenges, we adopt time-frequency signal processing based on fractional Fourier transform (FrFT) to achieve the unified representation of impairments, while employing a Transformer based neural networks (NN) to break through network performance limitations. To verify the effectiveness of the proposed estimation method, the numerical simulation is carried on a 5-channel polarization-division-multiplexed quadrature phase shift keying (PDM-QPSK) long haul optical transmission system with the symbol rate of 50 GBaud per channel, the mean absolute error (MAE) for SNRNL, OSNR, CD, and DGD estimation is 0.091 dB, 0.058 dB, 117 ps/nm, and 0.38 ps, and the monitoring window ranges from 0~20 dB, 10~30 dB, 0~51000 ps/nm, and 0~100 ps, respectively. Our proposed method achieves accurate estimation of linear and nonlinear impairments over a broad range, representing a significant advancement in the field of optical performance monitoring (OPM).

eess.SP↗
arXiv2023

When MiniBatch SGD Meets SplitFed Learning:Convergence Analysis and Performance Evaluation

Federated learning (FL) enables collaborative model training across distributed clients (e.g., edge devices) without sharing raw data. Yet, FL can be computationally expensive as the clients need to train the entire model multiple times. SplitFed learning (SFL) is a recent distributed approach that alleviates computation workload at the client device by splitting the model at a cut layer into two parts, where clients only need to train part of the model. However, SFL still suffers from the \textit{client drift} problem when clients' data are highly non-IID. To address this issue, we propose MiniBatch-SFL. This algorithm incorporates MiniBatch SGD into SFL, where the clients train the client-side model in an FL fashion while the server trains the server-side model similar to MiniBatch SGD. We analyze the convergence of MiniBatch-SFL and show that the bound of the expected loss can be obtained by analyzing the expected server-side and client-side model updates, respectively. The server-side updates do not depend on the non-IID degree of the clients' datasets and can potentially mitigate client drift. However, the client-side model relies on the non-IID degree and can be optimized by properly choosing the cut layer. Perhaps counter-intuitive, our empirical result shows that a latter position of the cut layer leads to a smaller average gradient divergence and a better algorithm performance. Moreover, numerical results show that MiniBatch-SFL achieves higher accuracy than conventional SFL and FL. The accuracy improvement can be up to 24.1\% and 17.1\% with highly non-IID data, respectively.

cs.LG↗
arXiv2023

Real-time FPGA Implementation of CNN-based Distributed Fiber Optic Vibration Event Recognition Method

Utilizing optical fibers to detect and pinpoint vibrations, Distributed Optical Fiber Vibration Sensing (DVS) technology provides real-time monitoring and surveillance of wide-reaching areas. This field has been leveraging Convolutional Neural Networks (CNN). Recently, a study has accomplished end-to-end vibration event recognition, enabling utilization of CNN-based DVS algorithms as real-time embedded system for edge computing in practical application situations. Considering the power consumption of central processing unit (CPU) and graphics processing unit (GPU), and the inflexibility of application-specific integrated circuit (ASIC), field-Programmable gate array (FPGA) is the optimal computing platform for the system. This paper proposes to compress pre-trained network and adopt a novel hardware structure, to design a fully on-chip, pipelined inference accelerator for CNN-based DVS algorithm, without fine tuning or re-training. This design allows for real-time processing with low power consumption and system requirement.An examination has been executed on an existing DVS algorithm based on a 40-layer CNN model comprising 2.7 million parameters. It is completely implemented on-chip, pipelined, with no reduction in accuracy.

eess.SP↗
arXiv2023

New Covert and Side Channels Based on Retirement

Intel processors utilize the retirement to orderly retire the micro-ops that have been executed out of order. To enhance retirement utilization, the retirement is dynamically shared between two logical cores on the same physical core. However, this shared retirement mechanism creates a potential vulnerability wherein an attacker can exploit the competition for retirement to infer the data of a victim on another logical core on the same physical core. Based on this leakage, we propose two new covert channels: the Different Instructions (DI) covert channel using different instructions for information transmission, and the Same Instructions (SI) covert channel using the same instructions to transmit information. The DI covert channel can achieve 98.5% accuracy with a bandwidth of 1450 Kbps, while the SI covert channel can achieve 94.85% accuracy with a bandwidth of 483.33 Kbps. Furthermore, this paper explores additional applications of retirement: Firstly, retirement is applied to Spectre attacks, resulting in a new variant of Spectre v1, which can achieve 94.17% accuracy with a bandwidth of 29 Kbps; Secondly, retirement is leveraged to infer the programs being executed by the victim, which can infer 10 integer benchmarks of SPEC with 89.28% accuracy. Finally, we discuss possible protection against new covert channels.

cs.CR↗