arXiv subjects

Xiaolong Zheng

Publications and source records attributed to Xiaolong Zheng.

32 records · Page 2Linked to original sources

arXiv2025

Leveraging Failed Samples: A Few-Shot and Training-Free Framework for Generalized Deepfake Detection

Recent deepfake detection studies often treat unseen sample detection as a ``zero-shot" task, training on images generated by known models but generalizing to unknown ones. A key real-world challenge arises when a model performs poorly on unknown samples, yet these samples remain available for analysis. This highlights that it should be approached as a ``few-shot" task, where effectively utilizing a small number of samples can lead to significant improvement. Unlike typical few-shot tasks focused on semantic understanding, deepfake detection prioritizes image realism, which closely mirrors real-world distributions. In this work, we propose the Few-shot Training-free Network (FTNet) for real-world few-shot deepfake detection. Simple yet effective, FTNet differs from traditional methods that rely on large-scale known data for training. Instead, FTNet uses only one fake samplefrom an evaluation set, mimicking the scenario where new samples emerge in the real world and can be gathered for use, without any training or parameter updates. During evaluation, each test sample is compared to the known fake and real samples, and it is classified based on the category of the nearest sample. We conduct a comprehensive analysis of AI-generated images from 29 different generative models and achieve a new SoTA performance, with an average improvement of 8.7\% compared to existing methods. This work introduces a fresh perspective on real-world deepfake detection: when the model struggles to generalize on a few-shot sample, leveraging the failed samples leads to better performance.

cs.CV↗

arXiv2025

VisualTrans: A Benchmark for Real-World Visual Transformation Reasoning

Visual transformation reasoning (VTR) is a vital cognitive capability that empowers intelligent agents to understand dynamic scenes, model causal relationships, and predict future states, and thereby guiding actions and laying the foundation for advanced intelligent systems. However, existing benchmarks suffer from a sim-to-real gap, limited task complexity, and incomplete reasoning coverage, limiting their practical use in real-world scenarios. To address these limitations, we introduce VisualTrans, the first comprehensive benchmark specifically designed for VTR in real-world human-object interaction scenarios. VisualTrans encompasses 12 semantically diverse manipulation tasks and systematically evaluates three essential reasoning dimensions - spatial, procedural, and quantitative - through 6 well-defined subtask types. The benchmark features 472 high-quality question-answer pairs in various formats, including multiple-choice, open-ended counting, and target enumeration. We introduce a scalable data construction pipeline built upon first-person manipulation videos, which integrates task selection, image pair extraction, automated metadata annotation with large multimodal models, and structured question generation. Human verification ensures the final benchmark is both high-quality and interpretable. Evaluations of various state-of-the-art vision-language models show strong performance in static spatial tasks. However, they reveal notable shortcomings in dynamic, multi-step reasoning scenarios, particularly in areas like intermediate state recognition and transformation sequence planning. These findings highlight fundamental weaknesses in temporal modeling and causal reasoning, providing clear directions for future research aimed at developing more capable and generalizable VTR systems. The dataset and code are available at https://github.com/WangYipu2002/VisualTrans.

cs.CV↗

arXiv2025

CRC-SGAD: Conformal Risk Control for Supervised Graph Anomaly Detection

Graph Anomaly Detection (GAD) is critical in security-sensitive domains, yet faces reliability challenges: miscalibrated confidence estimation (underconfidence in normal nodes, overconfidence in anomalies), adversarial vulnerability of derived confidence score under structural perturbations, and limited efficacy of conventional calibration methods for sparse anomaly patterns. Thus we propose CRC-SGAD, a framework integrating statistical risk control into GAD via two innovations: (1) A Dual-Threshold Conformal Risk Control mechanism that provides theoretically guaranteed bounds for both False Negative Rate (FNR) and False Positive Rate (FPR) through providing prediction sets; (2) A Subgraph-aware Spectral Graph Neural Calibrator (SSGNC) that optimizes node representations through adaptive spectral filtering while reducing the size of prediction sets via hybrid loss optimization. Experiments on four datasets and five GAD models demonstrate statistically significant improvements in FNR and FPR control and prediction set size. CRC-SGAD establishes a paradigm for statistically rigorous anomaly detection in graph-structured security applications.

cs.LG↗

arXiv2025

Alleviating Performance Disparity in Adversarial Spatiotemporal Graph Learning Under Zero-Inflated Distribution

Spatiotemporal Graph Learning (SGL) under Zero-Inflated Distribution (ZID) is crucial for urban risk management tasks, including crime prediction and traffic accident profiling. However, SGL models are vulnerable to adversarial attacks, compromising their practical utility. While adversarial training (AT) has been widely used to bolster model robustness, our study finds that traditional AT exacerbates performance disparities between majority and minority classes under ZID, potentially leading to irreparable losses due to underreporting critical risk events. In this paper, we first demonstrate the smaller top-k gradients and lower separability of minority class are key factors contributing to this disparity. To address these issues, we propose MinGRE, a framework for Minority Class Gradients and Representations Enhancement. MinGRE employs a multi-dimensional attention mechanism to reweight spatiotemporal gradients, minimizing the gradient distribution discrepancies across classes. Additionally, we introduce an uncertainty-guided contrastive loss to improve the inter-class separability and intra-class compactness of minority representations with higher uncertainty. Extensive experiments demonstrate that the MinGRE framework not only significantly reduces the performance disparity across classes but also achieves enhanced robustness compared to existing baselines. These findings underscore the potential of our method in fostering the development of more equitable and robust models.

cs.LG↗

arXiv2025

RoboBrain: A Unified Brain Model for Robotic Manipulation from Abstract to Concrete

Recent advancements in Multimodal Large Language Models (MLLMs) have shown remarkable capabilities across various multimodal contexts. However, their application in robotic scenarios, particularly for long-horizon manipulation tasks, reveals significant limitations. These limitations arise from the current MLLMs lacking three essential robotic brain capabilities: Planning Capability, which involves decomposing complex manipulation instructions into manageable sub-tasks; Affordance Perception, the ability to recognize and interpret the affordances of interactive objects; and Trajectory Prediction, the foresight to anticipate the complete manipulation trajectory necessary for successful execution. To enhance the robotic brain's core capabilities from abstract to concrete, we introduce ShareRobot, a high-quality heterogeneous dataset that labels multi-dimensional information such as task planning, object affordance, and end-effector trajectory. ShareRobot's diversity and accuracy have been meticulously refined by three human annotators. Building on this dataset, we developed RoboBrain, an MLLM-based model that combines robotic and general multi-modal data, utilizes a multi-stage training strategy, and incorporates long videos and high-resolution images to improve its robotic manipulation capabilities. Extensive experiments demonstrate that RoboBrain achieves state-of-the-art performance across various robotic tasks, highlighting its potential to advance robotic brain capabilities.

cs.RO↗

arXiv2025

Enhancing Adversarial Robustness of Vision-Language Models through Low-Rank Adaptation

Vision-Language Models (VLMs) play a crucial role in the advancement of Artificial General Intelligence (AGI). As AGI rapidly evolves, addressing security concerns has emerged as one of the most significant challenges for VLMs. In this paper, we present extensive experiments that expose the vulnerabilities of conventional adaptation methods for VLMs, highlighting significant security risks. Moreover, as VLMs grow in size, the application of traditional adversarial adaptation techniques incurs substantial computational costs. To address these issues, we propose a parameter-efficient adversarial adaptation method called \textbf{\textit{AdvLoRA}} based on Low-Rank Adaptation. We investigate and reveal the inherent low-rank properties involved in adversarial adaptation for VLMs. Different from LoRA, we enhance the efficiency and robustness of adversarial adaptation by introducing a novel reparameterization method that leverages parameter clustering and alignment. Additionally, we propose an adaptive parameter update strategy to further bolster robustness. These innovations enable our AdvLoRA to mitigate issues related to model security and resource wastage. Extensive experiments confirm the effectiveness and efficiency of AdvLoRA.

cs.CV↗

arXiv2024

Deep Causal Learning: Representation, Discovery and Inference

Causal learning has garnered significant attention in recent years because it reveals the essential relationships that underpin phenomena and delineates the mechanisms by which the world evolves. Nevertheless, traditional causal learning methods face numerous challenges and limitations, including high-dimensional, unstructured variables, combinatorial optimization problems, unobserved confounders, selection biases, and estimation inaccuracies. Deep causal learning, which leverages deep neural networks, offers innovative insights and solutions for addressing these challenges. Although numerous deep learning-based methods for causal discovery and inference have been proposed, there remains a dearth of reviews examining the underlying mechanisms by which deep learning can enhance causal learning. In this article, we comprehensively review how deep learning can contribute to causal learning by tackling traditional challenges across three key dimensions: representation, discovery, and inference. We emphasize that deep causal learning is pivotal for advancing the theoretical frontiers and broadening the practical applications of causal science. We conclude by summarizing open issues and outlining potential directions for future research.

cs.LG↗

arXiv2022

Adaptive Neural Network Backstepping Control Method for Aerial Manipulator Based on Variable Inertia Parameter Modeling

For the aerial manipulator that performs aerial work tasks, the actual operating environment it faces is very complex, and it is affected by internal and external multi-source disturbances. In this paper, to effectively improve the anti-disturbance control performance of the aerial manipulator, an adaptive neural network backstepping control method based on variable inertia parameter modeling is proposed. Firstly, for the intense internal coupling disturbance, we analyze and model it from the perspective of the generation mechanism of the coupling disturbance, and derive the dynamics model of the aerial manipulator system and the coupling disturbance model based on the variable inertia parameters. Through the proposed coupling disturbance model, we can compensate the strong coupling disturbance in a way of feedforward. Then, the adaptive neural network is proposed and applid to estimate and compensate the additional disturbances, and the closed-loop controller is designed based on the backstepping control method. Finally, we verify the correctness of the proposed coupling disturbance model through physical experiment under a large range motion of the manipulator. Two sets of comparative simulation results also prove the accurate estimation of the proposed adaptive neural network for additional disturbances and the effectiveness and superiority of the proposed control method.

cs.RO↗

arXiv2022

Crocs: Cross-Technology Clock Synchronization for WiFi and ZigBee

Clock synchronization is a key function in embedded wireless systems and networks. This issue is equally important and more challenging in IoT systems nowadays, which often include heterogeneous wireless devices that follow different wireless standards. Conventional solutions to this problem employ gateway-based indirect synchronization, which suffers low accuracy. This paper for the first time studies the problem of cross-technology clock synchronization. Our proposal called Crocs synchronizes WiFi and ZigBee devices by direct cross-technology communication. Crocs decouples the synchronization signal from the transmission of a timestamp. By incorporating a barker-code based beacon for time alignment and cross-technology transmission of timestamps, Crocs achieves robust and accurate synchronization among WiFi and ZigBee devices, with the synchronization error lower than 1 millisecond. We further make attempts to implement different cross-technology communication methods in Crocs and provide insight findings with regard to the achievable accuracy and expected overhead.

cs.NI↗

arXiv2022

AdaComm: Tracing Channel Dynamics for Reliable Cross-Technology Communication

Cross-Technology Communication (CTC) is an emerging technology to support direct communication between wireless devices that follow different standards. In spite of the many different proposals from the community to enable CTC, the performance aspect of CTC is an equally important problem but has seldom been studied before. We find this problem is extremely challenging, due to the following reasons: on one hand, a link for CTC is essentially different from a conventional wireless link. The conventional link indicators like RSSI (received signal strength indicator) and SNR (signal to noise ratio) cannot be used to directly characterize a CTC link. On the other hand, the indirect indicators like PER (packet error rate), which is adopted by many existing CTC proposals, cannot capture the short-term link behavior. As a result, the existing CTC proposals fail to keep reliable performance under dynamic channel conditions. In order to address the above challenge, we in this paper propose AdaComm, a generic framework to achieve self-adaptive CTC in dynamic channels. Instead of reactively adjusting the CTC sender, AdaComm adopts online learning mechanism to adaptively adjust the decoding model at the CTC receiver. The self-adaptive decoding model automatically learns the effective features directly from the raw received signals that are embedded with the current channel state. With the lossless channel information, AdaComm further adopts the fine tuning and full training modes to cope with the continuous and abrupt channel dynamics. We implement AdaComm and integrate it with two existing CTC approaches that respectively employ CSI (channel state information) and RSSI as the information carrier. The evaluation results demonstrate that AdaComm can significantly reduce the SER (symbol error rate) by 72.9% and 49.2%, respectively, compared with the existing approaches.

cs.NI↗

arXiv2022

FoVR: Attention-based VR Streaming through Bandwidth-limited Wireless Networks

Consumer Virtual Reality (VR) has been widely used in various application areas, such as entertainment and medicine. In spite of the superb immersion experience, to enable high-quality VR on untethered mobile devices remains an extremely challenging task. The high bandwidth demands of VR streaming generally overburden a conventional wireless connection, which affects the user experience and in turn limits the usability of VR in practice. In this paper, we propose FoVR, attention-based hierarchical VR streaming through bandwidth-limited wireless networks. The design of FoVR stems from the insight that human's vision is hierarchical, so that different areas in the field of view (FoV) can be served with VR content of different qualities. By exploiting the gaze tracking capacity of the VR devices, FoVR is able to accurately predict the user's attention so that the streaming of hierarchical VR can be appropriately scheduled. In this way, FoVR significantly reduces the bandwidth cost and computing cost while keeping high quality of user experience. We implement FoVR on a commercial VR device and evaluate its performance in various scenarios. The experiment results show that FoVR reduces the bandwidth cost by 88.9% and 76.2%, respectively compared to the original VR streaming and the state-of-the-art approach.

cs.NI↗

arXiv2022

Cross-Technology Communication for the Internet of Things: A Survey

The ever-developing Internet of Things (IoT) brings the prosperity of wireless sensing and control applications. In many scenarios, different wireless technologies coexist in the shared frequency medium as well as the physical space. Such wireless coexistence may lead to serious cross-technology interference (CTI) problems, e.g. channel competition, signal collision, throughput degradation. Compared with traditional methods like interference avoidance, tolerance, and concurrency mechanism, direct and timely information exchange among heterogeneous devices is therefore a fundamental requirement to ensure the usability, inter-operability, and reliability of the IoT. Under this circumstance, Cross-Technology Communication (CTC) technique thus becomes a hot topic in both academic and industrial fields, which aims at directly exchanging data among heterogeneous devices that follow different standards. This paper comprehensively summarizes the CTC techniques and reveals that the key challenge for CTC lies in the heterogeneity of IoT devices, including the incompatibility of technical standards and the asymmetry of connection capability. Based on the above finding, we present a taxonomy of the existing CTC works (packet-level CTCs and physical-level CTCs) and compare the existing CTC techniques in terms of throughput, reliability, hardware modification, and concurrency.

cs.NI↗

arXiv2021

Self-adaptive Multi-task Particle Swarm Optimization

Multi-task optimization (MTO) studies how to simultaneously solve multiple optimization problems for the purpose of obtaining better performance on each problem. Over the past few years, evolutionary MTO (EMTO) was proposed to handle MTO problems via evolutionary algorithms. So far, many EMTO algorithms have been developed and demonstrated well performance on solving real-world problems. However, there remain many works to do in adapting knowledge transfer to task relatedness in EMTO. Different from the existing works, we develop a self-adaptive multi-task particle swarm optimization (SaMTPSO) through the developed knowledge transfer adaptation strategy, the focus search strategy and the knowledge incorporation strategy. In the knowledge transfer adaptation strategy, each task has a knowledge source pool that consists of all knowledge sources. Each source (task) outputs knowledge to the task. And knowledge transfer adapts to task relatedness via individuals' choice on different sources of a pool, where the chosen probabilities for different sources are computed respectively according to task's success rate in generating improved solutions via these sources. In the focus search strategy, if there is no knowledge source benefit the optimization of a task, then all knowledge sources in the task's pool are forbidden to be utilized except the task, which helps to improve the performance of the proposed algorithm. Note that the task itself is as a knowledge source of its own. In the knowledge incorporation strategy, two different forms are developed to help the SaMTPSO explore and exploit the transferred knowledge from a chosen source, each leading to a version of the SaMTPSO. Several experiments are conducted on two test suites. The results of the SaMTPSO are comparing to that of 3 popular EMTO algorithms and a particle swarm algorithm, which demonstrates the superiority of the SaMTPSO.

cs.NE↗

arXiv2019

Simultaneous polarization transformation and amplification of multi-petawatt laser pulses in magnetized plasmas

With increasing laser peak power, the generation and manipulation of high-power laser pulses becomes a growing challenge for conventional solid-state optics due to their limited damage threshold. As a result, plasma-based optical components which can sustain extremely high fields are attracting increasing interest. Here, we propose a type of plasma waveplate based on magneto-optical birefringence under a transverse magnetic field, which can work under extremely high laser power. Importantly, this waveplate can simultaneously alter the polarization state and boost the peak laser power. It is demonstrated numerically that an initially linearly polarized laser pulse with 5 petawatt peak power can be converted into a circularly polarized pulse with a peak power higher than 10 petawatts by such a waveplate with a centimeter-scale diameter. The energy conversion efficiency of the polarization transformation is about $98\%$. The necessary waveplate thickness is shown to scale inversely with plasma electron density $n_e$ and the square of magnetic field $B_0$, and it is about 1 cm for $n_e=3\times 10^{20}$ cm$^{-3}$ and $B_0=100$ T. The proposed plasma waveplate and other plasma-based optical components can play a critical role for the effective utilization of multi-petawatt laser systems.

physics.plasm-ph↗