arXiv subjects

Peng Liu

Publications and source records attributed to Peng Liu.

At least 217 records · Page 12Linked to original sources

arXiv2024

Identifying Non-Control Security-Critical Data through Program Dependence Learning

As control-flow protection gets widely deployed, it is difficult for attackers to corrupt control-data and achieve control-flow hijacking. Instead, data-oriented attacks, which manipulate non-control data, have been demonstrated to be feasible and powerful. In data-oriented attacks, a fundamental step is to identify non-control, security-critical data. However, critical data identification processes are not scalable in previous works, because they mainly rely on tedious human efforts to identify critical data. To address this issue, we propose a novel approach that combines traditional program analysis with deep learning. At a higher level, by examining how analysts identify critical data, we first propose dynamic analysis algorithms to identify the program semantics (and features) that are correlated with the impact of a critical data. Then, motivated by the unique challenges in the critical data identification task, we formalize the distinguishing features and use customized program dependence graphs (PDG) to embed the features. Different from previous works using deep learning to learn basic program semantics, this paper adopts a special neural network architecture that can capture the long dependency paths (in the PDG), through which a critical variable propagates its impact. We have implemented a fully-automatic toolchain and conducted comprehensive evaluations. According to the evaluations, our model can achieve 90% accuracy. The toolchain uncovers 80 potential critical variables in Google FuzzBench. In addition, we demonstrate the harmfulness of the exploits using the identified critical variables by simulating 7 data-oriented attacks through GDB.

cs.CR↗

arXiv2024

Hidden You Malicious Goal Into Benign Narratives: Jailbreak Large Language Models through Logic Chain Injection

Jailbreak attacks on Language Model Models (LLMs) entail crafting prompts aimed at exploiting the models to generate malicious content. Existing jailbreak attacks can successfully deceive the LLMs, however they cannot deceive the human. This paper proposes a new type of jailbreak attacks which can deceive both the LLMs and human (i.e., security analyst). The key insight of our idea is borrowed from the social psychology - that is human are easily deceived if the lie is hidden in truth. Based on this insight, we proposed the logic-chain injection attacks to inject malicious intention into benign truth. Logic-chain injection attack firstly dissembles its malicious target into a chain of benign narrations, and then distribute narrations into a related benign article, with undoubted facts. In this way, newly generate prompt cannot only deceive the LLMs, but also deceive human.

cs.CR↗

arXiv2024

LoopAnimate: Loopable Salient Object Animation

Research on diffusion model-based video generation has advanced rapidly. However, limitations in object fidelity and generation length hinder its practical applications. Additionally, specific domains like animated wallpapers require seamless looping, where the first and last frames of the video match seamlessly. To address these challenges, this paper proposes LoopAnimate, a novel method for generating videos with consistent start and end frames. To enhance object fidelity, we introduce a framework that decouples multi-level image appearance and textual semantic information. Building upon an image-to-image diffusion model, our approach incorporates both pixel-level and feature-level information from the input image, injecting image appearance and textual semantic embeddings at different positions of the diffusion model. Existing UNet-based video generation models require to input the entire videos during training to encode temporal and positional information at once. However, due to limitations in GPU memory, the number of frames is typically restricted to 16. To address this, this paper proposes a three-stage training strategy with progressively increasing frame numbers and reducing fine-tuning modules. Additionally, we introduce the Temporal E nhanced Motion Module(TEMM) to extend the capacity for encoding temporal and positional information up to 36 frames. The proposed LoopAnimate, which for the first time extends the single-pass generation length of UNet-based video generation models to 35 frames while maintaining high-quality video generation. Experiments demonstrate that LoopAnimate achieves state-of-the-art performance in both objective metrics, such as fidelity and temporal consistency, and subjective evaluation results.

cs.CV↗

arXiv2024

Factor risk measures

This paper introduces and studies factor risk measures. While risk measures only rely on the distribution of a loss random variable, in many cases risk needs to be measured relative to some major factors. In this paper, we introduce a double-argument mapping as a risk measure to assess the risk relative to a vector of factors, called factor risk measure. The factor risk measure only depends on the joint distribution of the risk and the factors. A set of natural axioms are discussed, and particularly distortion, quantile, linear and coherent factor risk measures are introduced and characterized. Moreover, we introduce a large set of concrete factor risk measures and many of them are new to the literature, which are interpreted in the context of regulatory capital requirement. Finally, the distortion factor risk measures are applied in the risk-sharing problem and some numerical examples are presented to show the difference between the Value-at-Risk and the quantile factor risk measures.

q-fin.MF↗

arXiv2024

AUEditNet: Dual-Branch Facial Action Unit Intensity Manipulation with Implicit Disentanglement

Facial action unit (AU) intensity plays a pivotal role in quantifying fine-grained expression behaviors, which is an effective condition for facial expression manipulation. However, publicly available datasets containing intensity annotations for multiple AUs remain severely limited, often featuring a restricted number of subjects. This limitation places challenges to the AU intensity manipulation in images due to disentanglement issues, leading researchers to resort to other large datasets with pretrained AU intensity estimators for pseudo labels. In addressing this constraint and fully leveraging manual annotations of AU intensities for precise manipulation, we introduce AUEditNet. Our proposed model achieves impressive intensity manipulation across 12 AUs, trained effectively with only 18 subjects. Utilizing a dual-branch architecture, our approach achieves comprehensive disentanglement of facial attributes and identity without necessitating additional loss functions or implementing with large batch sizes. This approach offers a potential solution to achieve desired facial attribute editing despite the dataset's limited subject count. Our experiments demonstrate AUEditNet's superior accuracy in editing AU intensities, affirming its capability in disentangling facial attributes and identity within a limited subject pool. AUEditNet allows conditioning by either intensity values or target images, eliminating the need for constructing AU combinations for specific facial expression synthesis. Moreover, AU intensity estimation, as a downstream task, validates the consistency between real and edited images, confirming the effectiveness of our proposed AU intensity manipulation method.

cs.CV↗

arXiv2024

Probing Berry phase effect in topological surface states

We have observed the Berry phase effect associated with interband coherence in topological surface states (TSSs) using two-color high-harmonic spectroscopy. This Berry phase accumulates along the evolution path of strong field-driven election-hole quasiparticles in electronic bands with strong spin-orbit coupling. By introducing a secondary weak field, we perturb the evolution of Dirac fermions in TSSs and thus provide access to the Berry phase. We observe a significant shift in the oscillation phase of the even-order harmonics from the spectral interferogram. We reveal that such a modulation feature is linked to the geometric phase acquired in the nonperturbative dynamics of TSSs. Furthermore, we show that the overwhelming Berry phase effect can significantly deform the quantum paths of electron-hole pairs, thus enhancing the ability to harness electron spin using lightwaves in quantum materials with strong spin-orbit interactions.

physics.optics↗

arXiv2024

Modern Statistical Models and Methods for Estimating Fatigue-Life and Fatigue-Strength Distributions from Experimental Data

Engineers and scientists have been collecting and analyzing fatigue data since the 1800s to ensure the reliability of life-critical structures. Applications include (but are not limited to) bridges, building structures, aircraft and spacecraft components, ships, ground-based vehicles, and medical devices. Engineers need to estimate S-N relationships (Stress or Strain versus Number of cycles to failure), typically with a focus on estimating small quantiles of the fatigue-life distribution. Estimates from this kind of model are used as input to models (e.g., cumulative damage models) that predict failure-time distributions under varying stress patterns. Also, design engineers need to estimate lower-tail quantiles of the closely related fatigue-strength distribution. The history of applying incorrect statistical methods is nearly as long and such practices continue to the present. Examples include treating the applied stress (or strain) as the response and the number of cycles to failure as the explanatory variable in regression analyses (because of the need to estimate strength distributions) and ignoring or otherwise mishandling censored observations (known as runouts in the fatigue literature). The first part of the paper reviews the traditional modeling approach where a fatigue-life model is specified. We then show how this specification induces a corresponding fatigue-strength model. The second part of the paper presents a novel alternative modeling approach where a fatigue-strength model is specified and a corresponding fatigue-life model is induced. We explain and illustrate the important advantages of this new modeling approach.

stat.ME↗

arXiv2024

Variational Dynamic for Self-Supervised Exploration in Deep Reinforcement Learning

Efficient exploration remains a challenging problem in reinforcement learning, especially for tasks where extrinsic rewards from environments are sparse or even totally disregarded. Significant advances based on intrinsic motivation show promising results in simple environments but often get stuck in environments with multimodal and stochastic dynamics. In this work, we propose a variational dynamic model based on the conditional variational inference to model the multimodality and stochasticity. We consider the environmental state-action transition as a conditional generative process by generating the next-state prediction under the condition of the current state, action, and latent variable, which provides a better understanding of the dynamics and leads a better performance in exploration. We derive an upper bound of the negative log-likelihood of the environmental transition and use such an upper bound as the intrinsic reward for exploration, which allows the agent to learn skills by self-supervised exploration without observing extrinsic rewards. We evaluate the proposed method on several image-based simulation tasks and a real robotic manipulating task. Our method outperforms several state-of-the-art environment model-based exploration approaches.

cs.LG↗

arXiv2024

A duality between utility transforms and probability distortions

In this paper, we establish a mathematical duality between utility transforms and probability distortions. These transforms play a central role in decision under risk by forming the foundation for the classic theories of expected utility, dual utility, and rank-dependent utility. Our main results establish that probability distortions are characterized by commutation with utility transforms, and utility transforms are characterized by commutation with probability distortions. These results require no additional conditions, and hence each class can be axiomatized with only one property. Moreover, under monotonicity, rank-dependent utility transforms can be characterized by set commutation with either utility transforms or probability distortions.

econ.TH↗

arXiv2024

On Equivalence of Likelihood-Based Confidence Bands for Fatigue-Life and Fatigue-Strength Distributions

Fatigue data arise in many research and applied areas and there have been statistical methods developed to model and analyze such data. The distributions of fatigue life and fatigue strength are often of interest to engineers designing products that might fail due to fatigue from cyclic-stress loading. Based on a specified statistical model and the maximum likelihood method, the cumulative distribution function (cdf) and quantile function (qf) can be estimated for the fatigue-life and fatigue-strength distributions. Likelihood-based confidence bands then can be obtained for the cdf and qf. This paper provides equivalence results for confidence bands for fatigue-life and fatigue-strength models. These results are useful for data analysis and computing implementation. We show (a) the equivalence of the confidence bands for the fatigue-life cdf and the fatigue-life qf, (b) the equivalence of confidence bands for the fatigue-strength cdf and the fatigue-strength qf, and (c) the equivalence of confidence bands for the fatigue-life qf and the fatigue-strength qf. Then we illustrate the usefulness of those equivalence results with two examples using experimental fatigue data.

stat.ME↗

arXiv2024

Robust Numerical Methods for Nonlinear Regression

Many scientific and engineering applications require fitting regression models that are nonlinear in the parameters. Advances in computer hardware and software in recent decades have made it easier to fit such models. Relative to fitting regression models that are linear in the parameters, however, fitting nonlinear regression models is more complicated. In particular, software like the $\texttt{nls}$ R function requires care in how the model is parameterized and how initial values are chosen for the maximum likelihood iterations. Often special diagnostics are needed to detect and suggest approaches for dealing with identifiability problems that can arise with such model fitting. When using Bayesian inference, there is the added complication of having to specify (often noninformative or weakly informative) prior distributions. Generally, the details for these tasks must be determined for each new nonlinear regression model. This paper provides a step-by-step procedure for specifying these details for any appropriate nonlinear regression model. Following the procedure will result in a numerically robust algorithm for fitting the nonlinear regression model. We illustrate the methods with three different nonlinear models that are used in the analysis of experimental fatigue data and we include two detailed numerical examples.

stat.ME↗

arXiv2024

DeepCRE: Transforming Drug R&D via AI-Driven Cross-drug Response Evaluation

The fields of therapeutic application and drug research and development (R&D) both face substantial challenges, i.e., the therapeutic domain calls for more treatment alternatives, while numerous promising pre-clinical drugs have failed in clinical trials. One of the reasons is the inadequacy of Cross-drug Response Evaluation (CRE) during the late stages of drug R&D. Although in-silico CRE models bring a promising solution, existing methodologies are restricted to early stages of drug R&D, such as target and cell-line levels, offering limited improvement to clinical success rates. Herein, we introduce DeepCRE, a pioneering AI model designed to predict CRE effectively in the late stages of drug R&D. DeepCRE outperforms the existing best models by achieving an average performance improvement of 17.7% in patient-level CRE, and a 5-fold increase in indication-level CRE, facilitating more accurate personalized treatment predictions and better pharmaceutical value assessment for indications, respectively. Furthermore, DeepCRE has identified a set of six drug candidates that show significantly greater effectiveness than a comparator set of two approved drugs in 5/8 colorectal cancer organoids. This demonstrates the capability of DeepCRE to systematically uncover a spectrum of drug candidates with enhanced therapeutic effects, highlighting its potential to transform drug R&D.

cs.AI↗

arXiv2024

Dispersive-wave-agile optical frequency division

The remarkable frequency stability of resonant systems in the optical domain (optical cavities and atomic transitions) can be harnessed at frequency scales accessible by electronics using optical frequency division. This capability is revolutionizing technologies spanning time keeping to high-performance electrical signal sources. A version of the technique called 2-point optical frequency division (2P-OFD) is proving advantageous for application to high-performance signal sources. In 2P-OFD, an optical cavity anchors two spectral endpoints defined by lines of a frequency comb. The comb need not be self-referenced, which greatly simplifies the system architecture and reduces power requirements. Here, a 2P-OFD microwave signal source is demonstrated with record-low phase noise using a microcomb. Key to this advance is a spectral endpoint defined by a frequency agile single-mode dispersive wave that is emitted by the microcomb soliton. Moreover, the system frequency reference is a compact all-solid-state optical cavity with a record Q-factor. The results advance integrable microcomb-based signal sources into the performance realm of much larger microwave sources.

physics.optics↗

arXiv2024

OmDet: Large-scale vision-language multi-dataset pre-training with multimodal detection network

The advancement of object detection (OD) in open-vocabulary and open-world scenarios is a critical challenge in computer vision. This work introduces OmDet, a novel language-aware object detection architecture, and an innovative training mechanism that harnesses continual learning and multi-dataset vision-language pre-training. Leveraging natural language as a universal knowledge representation, OmDet accumulates a "visual vocabulary" from diverse datasets, unifying the task as a language-conditioned detection framework. Our multimodal detection network (MDN) overcomes the challenges of multi-dataset joint training and generalizes to numerous training datasets without manual label taxonomy merging. We demonstrate superior performance of OmDet over strong baselines in object detection in the wild, open-vocabulary detection, and phrase grounding, achieving state-of-the-art results. Ablation studies reveal the impact of scaling the pre-training visual vocabulary, indicating a promising direction for further expansion to larger datasets. The effectiveness of our deep fusion approach is underscored by its ability to learn jointly from multiple datasets, enhancing performance through knowledge sharing.

cs.CV↗

arXiv2024

Safeguarding the Truth of High-Value Price Oracle Task: A Dynamically Adjusted Truth Discovery Method

In recent years, the Decentralized Finance (DeFi) market has witnessed numerous attacks on the price oracle, leading to substantial economic losses. Despite the advent of truth discovery methods opening up new avenues for oracle development, it falls short in addressing high-value attacks on price oracle tasks. Consequently, this paper introduces a dynamically adjusted truth discovery method safeguarding the truth of high-value price oracle tasks. In the truth aggregation stage, we enhance future considerations to improve the precision of aggregated truth. During the credibility update phase, credibility is dynamically assessed based on the task's value and the Cumulative Potential Economic Contribution (CPEC) of information sources. Experimental results demonstrate a significant reduction in data deviation by 65.8\% and potential economic loss by 66.5\%, compared to the baseline scheme, in the presence of high-value attacks.

cs.GT↗

arXiv2024

Exact non-Hermitian mobility edges and robust flat bands in two-dimensional Lieb lattices with imaginary quasiperiodic potentials

The mobility edge (ME) is a critical energy delineates the boundary between extended and localized states within the energy spectrum, and it plays a crucial role in understanding the metal-insulator transition in disordered or quasiperiodic systems. While there have been extensive studies on MEs in one-dimensional non-Hermitian (NH) quasiperiodic lattices recently, the investigation of exact NH MEs in two-dimensional (2D) cases remains rare. In the present study, we introduce a 2D dissipative Lieb lattice (DLL) model with imaginary quasiperiodic potentials applied solely to the vertices of the Lieb lattice. By mapping this DLL model to the 2D NH Aubry-Andr{é}-Harper (AAH) model, we analytically derive the exact ME and find it associated with the absolute eigenenergies. We find that the eigenvalues of extended states are purely imaginary when the quasiperiodic potential is strong enough. Additionally, we demonstrate that the introduction of imaginary quasiperiodic potentials does not disrupt the flat bands inherent in the system. Finally, we propose a theoretical framework for realizing our model using the Lindblad master equation. Our results pave the way for further investigation of exact NH MEs and flat bands in 2D dissipative quasiperiodic systems.

cond-mat.dis-nn↗

arXiv2024

ROPNN: Detection of ROP Payloads Using Deep Neural Networks

Return-oriented programming (ROP) is a code reuse attack that chains short snippets of existing code to perform arbitrary operations on target machines. Existing detection methods against ROP exhibit unsatisfactory detection accuracy and/or have high runtime overhead. In this paper, we present ROPNN, which innovatively combines address space layout guided disassembly and deep neural networks to detect ROP payloads. The disassembler treats application input data as code pointers and aims to find any potential gadget chains, which are then classified by a deep neural network as benign or malicious. Our experiments show that ROPNN has high detection rate (99.3%) and a very low false positive rate (0.01%). ROPNN successfully detects all of the 100 real-world ROP exploits that are collected in-the-wild, created manually or created by ROP exploit generation tools. Additionally, ROPNN detects all 10 ROP exploits that can bypass Bin-CFI. ROPNN is non-intrusive and does not incur any runtime overhead to the protected program.

cs.CR↗

arXiv2024

Auxiliary Reward Generation with Transition Distance Representation Learning

Reinforcement learning (RL) has shown its strength in challenging sequential decision-making problems. The reward function in RL is crucial to the learning performance, as it serves as a measure of the task completion degree. In real-world problems, the rewards are predominantly human-designed, which requires laborious tuning, and is easily affected by human cognitive biases. To achieve automatic auxiliary reward generation, we propose a novel representation learning approach that can measure the ``transition distance'' between states. Building upon these representations, we introduce an auxiliary reward generation technique for both single-task and skill-chaining scenarios without the need for human knowledge. The proposed approach is evaluated in a wide range of manipulation tasks. The experiment results demonstrate the effectiveness of measuring the transition distance between states and the induced improvement by auxiliary rewards, which not only promotes better learning efficiency but also increases convergent stability.

cs.LG↗