Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Gérard MemmiSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works166 from public data
- 331
- 245
- 245
- DeepSweep: An Evaluation Framework for Mitigating DNN Backdoor Attacks using Data Augmentation230
A systematic approach is proposed to discover the optimal policies for defending against different backdoor attacks by comprehensively evaluating 71 state-of-the-art data augmentation functions and envision this framework can be a good benchmark tool to advance future DNN backdoor studies.
- 198
- 155
- Selective encryption on ECG data in body sensor network based on supervised machine learning155
This paper re-define the SE schemes in a practical scenario of securely outsourcing the electrocardiogram (ECG) data in the untrusted BSN environment and gives the SE design based on this classification use case to protect the ECG data against illegal classification at the attacker sides which protects the patients’ data privacy.
- 145
- A Dynamic Scalable Blockchain Based Communication Architecture for IoT137
A dynamic Blockchain based trust system is proposed to provide a dynamic and scalable communication architecture for IoT networks and a case study is presented to further discuss the security issues and provide future research directions.
- 135
- 126
- System Log Parsing: A Survey85
A comprehensive survey on log parsing is provided as a helpful resource for system administrators and domain experts to choose the most desirable open-source solution or implement new ones based on application-specific requirements.
- CLAP: Learning Transferable Binary Code Representations with Natural Language Supervision79
This work presents CLAP (Contrastive Language-Assembly Pre-training), which employs natural language supervision to learn better representations of binary code and get better transferability and proposes an efficient dataset engine that could automatically generate a large and diverse dataset comprising of binary code and corresponding natural language explanations.
- A Networking Perspective on Starlink's Self-Driving LEO Mega-Constellation75
This paper designs a domain-specific method to characterize key components in Starlink's autonomous driving from various public space situational awareness datasets, including continuous orbit maintenance, collision avoidance, and maneuvers between orbital shells.
- 70
- Privacy-preserving Collaborative Learning with Automatic Transformation Search67
This paper proposes to leverage data augmentation to defeat reconstruction attacks: by preprocessing sensitive images with carefully-selected transformation policies, it becomes infeasible for the adversary to extract any useful information from the corresponding gradients.
- A Data Augmentation-Based Defense Method Against Adversarial Attacks in Neural Networks67
This paper focuses on developing a lightweight defense method that can efficiently invalidate full whitebox adversarial attacks with the compatibility of real-life constraints and demonstrates outstanding robustness and efficiency.
- An efficient key distribution system for data fusion in V2X heterogeneous networks61
This paper proposes an efficient and practical data fusion trust system for the multi-source and multi-formats of data exchange in the V2X heterogeneous networks and presents a location-based PKI system with acceleration brought by General Purpose Graphic Processing Unit for efficient key distribution with a high level of trust achieved.
- All-Or-Nothing data protection for ubiquitous communication: Challenges and perspectives60
This paper collates research works across a number of topics in AON, from historical developments to the recent research proposal, and introduces new definitions and category methods to provide useful reference and perspectives for the state of AON research and development.
- NFV Platforms: Taxonomy, Design Choices and Future Challenges58
This study gives a detailed guideline for network operators or service providers to choose the most appropriate NFV platform based on their respective requirements and also provides guidelines for implementing new NFV platforms.
- Aegis: Mitigating Targeted Bit-flip Attacks against Deep Neural Networks55
A dynamic-exit mechanism to attach extra internal classifiers (ICs) to hidden layers and randomly selects ICs for predictions during each inference to significantly increase the attack cost for the adaptive attacks where all defense mechanisms are transparent to the adversary.
- 55
- Novel denial-of-service attacks against cloud-based multi-robot systems46
By analyzing different attack vectors in cloud-robotic platforms, this paper proposes three new DoS attacks, which manipulate the network resources, micro-architecture resources, and function parameters respectively, and alerts the robotics community to these catastrophic attacks.
- An Engorgio Prompt Makes Large Language Model Babble on44
This paper designs Engorgio, a novel methodology, to efficiently generate adversarial Engorgio prompts to affect the target LLM's service availability, and proposes novel loss functions to stably suppress the appearance of thetoken, whose occurrence will interrupt the LLM's generation process.
- 44
- 43
- One-bit Flip is All You Need: When Bit-flip Attack Meets Model Training42
A training-assisted bit flip attack is proposed, in which the adversary is involved in the training stage to build a high-risk model to release and this high-risk model, obtained coupled with a corresponding malicious model, behaves normally and can escape various detection methods.
- 42
- CEBin: A Cost-Effective Framework for Large-Scale Binary Code Similarity Detection41
A cost-effective BCSD framework, CEBin, which fuses embedding-based and comparison-based approaches to significantly improve accuracy while minimizing overheads and is able to provide an effective and efficient solution for detecting similar code (including vulnerable ones) in large-scale software ecosystems.
- Fine-tuning Is Not Enough: A Simple yet Effective Watermark Removal Attack for DNN Models41
A novel watermark removal attack by combining imperceptible pattern embedding and spatial-level transformations, which can effectively and blindly destroy the memorization of watermarked models to the watermark samples.
- VideoShield: Regulating Diffusion-based Video Generation Models via Watermarking39
VideoShield is a novel watermarking framework specifically designed for popular diffusion-based video generation models that effectively extracts watermarks and detects tamper without compromising video quality and is applicable to image generation models, enabling tamper detection in generated images as well.
- 39
- 38
- BET: black-box efficient testing for convolutional neural networks36
The results show that BET significantly outperforms existing white-box and black-box testing methods considering the effective error-inducing inputs found in a fixed query/inference budget, and shows that the error- inducing inputs found by BET can be used to fine-tune the target model, improving its accuracy by up to 3%.
- 36
- 27
- 27
- 25
- HAPE: A programmable big knowledge graph platform22
The current version of Heaven Ape is a prototype, which consists of three parts: a big knowledge graph knowledge base, a knowledge graph browser on the client side, and a knowledge graphs operating system on the server side.
- 21
- Watermarking Pre-trained Encoders in Contrastive Learning21
This work proposes the first watermarking methodology for the pre-trained encoders using a task-agnostic loss function to effectively embed into the encoder a backdoor as the watermark and indicates high effectiveness and robustness against different adversarial operations.
- 19
- 18
- 17
- 17
- 17
- Fast Selective Encryption Methods for Bitmap Images17
A new solution for image protection within resource environments offered by commodity computers such as desktops, laptops, tablets, or even smartphones is derived by parallelizing selective encryption and using available GPU Graphic Process Unit acceleration.
- PhyScout: Detecting Sensor Spoofing Attacks via Spatio-temporal Consistency16
Compared to existing defense solutions, PhyScout offers rapid identification of sensor attacks (within 100ms) with low performance overhead (CPU-based), and conflict visualization, and presents new avenues for future research in robust and efficient defense mechanisms against sensor spoofing attacks.
- 16
- 16
- Model Uncertainty Based Annotation Error Fixing for Web Attack Detection16
This work proposes model uncertainty to evaluate the prediction made by the DL based web attack model, and proves the efficiency of the detection model: with the aid of GPU acceleration, it is capable of tackling attacks on the fly.
- DC coefficient recovery for JPEG images in ubiquitous communication systems16
An image content recovery method for JPEG images that can recover the image content by estimating the DC coefficients without any pre-know knowledge is proposed so that fault tolerance can be achieved at the receivers’ ends for ubiquitous communication systems.
- Improving Adversarial Robustness of 3D Point Cloud Classification Models15
Two innovative methodologies to improve the adversarial robustness of 3D point cloud classification models are designed, including a novel point cloud architecture which can smooth and disrupt the adversarial perturbations and a novel data augmentation strategy to adaptively balance the model usability and robustness.
- 15
- 15
- What Makes a Good LLM Agent for Real-world Penetration Testing?14
Excalibur is presented, a penetration testing agent that couples strong tooling with difficulty-aware planning and addresses a limitation that model scaling alone does not eliminate, showing that difficulty-aware planning yields consistent end-to-end gains across models.
- 14
- 13
- 13
- 12
- Mind Your Heart: Stealthy Backdoor Attack on Dynamic Deep Neural Network in Edge Computing11
This paper proposes a novel backdoor attack specifically on the dynamic multi-exit DNN models by poisoning one DNN model’s shallow hidden layers targeting not this vanilla DNNmodel but only its dynamically deployed multi- exit architectures.
- The Hidden Vulnerability of Watermarking for Deep Neural Networks.11
A novel watermark removal attack is designed, which can defeat state-of-the-art solutions without any prior knowledge of the adopted watermarking technique and training samples, and a novel preprocessing function is proposed, which embeds imperceptible patterns and performs spatial-level transformations over the input.
- 11
- ShieldVLM: Safeguarding the Multimodal Implicit Toxicity via Deliberative Reasoning with LVLMs: ShieldVLM10
ShieldVLM is built, a model which identifies implicit toxicity in multimodal statements, prompts and dialogs via deliberative cross-modal reasoning and outperforms existing strong baselines in detecting both implicit and explicit toxicity.
- VisionGuard: Secure and Robust Visual Perception of Autonomous Vehicles in Practice10
The key of VisionGuard is to leverage the spatiotemporal inconsistency property of PAEs to detect anomalies and it predicts the motion states from historical ones and compares them with the current driving states to identify any motion inconsistency caused by physical attacks.
- 10
- 10
- DREAM: Scalable Red Teaming for Text-to-Image Generative Systems via Distribution Modeling9
DREAM is proposed, a scalable red teaming framework to automatically uncover diverse problematic prompts from a given T2I system, and GC-SPSA, an efficient optimization algorithm that provides stable gradient estimates through the long and potentially non-differentiable T2I pipeline.
- A Case for Application-Aware Space Radiation Tolerance in Orbital Computing9
This work makes a case for cost-effective space radiation tolerance using application domain knowledge and exploits the uneven SEU/MCU sensitivity across DNN layers and MCUs' spatial correlation for lightweight radiation-tolerant in-orbit AI computing.
- COSMIC: Compress Satellite Images Efficiently via Diffusion Compensation9
This paper designs a lightweight encoder and proposes COSMIC, a simple yet effective learned compression solution to transmit satellite images and shows that COSMIC outperforms state-of-the-art baselines on both perceptual and distortion metrics.
- 9
- 8
- 8
- 8
- 8
- 7
- 7
- Mitigating Query-based Neural Network Fingerprinting via Data Augmentation7
A novel attack to mitigate query-based fingerprinting methods based on data augmentation methods by proposing a randomized transformation on input samples to significantly mislead the fingerprint samples’ prediction and compromise the IP verification.
- 7
- 6
- 6
- 6
- 6
- State-Dependent Safety Failures in Multi-Turn Language Model Interaction5
This work introduces STAR, a state-oriented diagnostic framework that treats dialogue history as a state transition operator and enables controlled analysis of safety behavior along interaction trajectories and finds that systems that appear robust under static evaluation can undergo rapid and reproducible safety collapse under structured multi-turn interaction.
- Revisiting Backdoor Attacks on LLMs: A Stealthy and Practical Poisoning Framework via Harmless Inputs5
A novel clean-data backdoor attack for jailbreaking LLMs that overfits triggers to a fixed, benign-sounding positive reply prefix using harmless QA pairs, and employs a gradient-based coordinate optimization to enhance the universal trigger.
- FaceID-6M: A Large-Scale, Open-Source FaceID Customization Dataset5
The first large-scale, open-source FaceID dataset containing 6 million high-quality text-image pairs is collected and released, and it is demonstrated that models trained on the FaceID-6M dataset achieve performance that is comparable to, and slightly better than currently available industrial models.
- 5
- 5
- Towards Understanding the Cognitive Habits of Large Reasoning Models4
CogTest, a principled benchmark designed to evaluate large reasoning models’ cognitive habits, reveals that LRMs, unlike conventional LLMs, not only exhibit human-like habits but also adaptively deploy them according to different tasks and suggests that studying persistent behavioral patterns in LRMs’ CoTs is a valuable step toward deeper understanding of LLM misbehavior.
- SafeRedir: Prompt Embedding Redirection for Robust Unlearning in Image Generation Models4
SafeRedir is introduced, a lightweight inference-time framework for robust unlearning via prompt embedding redirection that achieves effective unlearning capability, high semantic and perceptual preservation, robust image quality, and enhanced resistance to adversarial attacks.
- 4
- BitHydra: Towards Bit-flip Inference Cost Attack against Large Language Models4
This work proposes BitHydra, a framework that addresses the unique optimization challenge of identifying the exact weight bits that maximize generation cost via the Alternating Direction Method of Multipliers (ADMM), and demonstrates the effectiveness of the ADMM-based formulation against both standard models and potential defenses.
- 4
- Laser Shield: a Physical Defense with Polarizer against Laser Attacks on Autonomous Driving Systems4
This work proposes Laser Shield which leverages a polarizer along with a min-energy rotation mechanism to eliminate adversarial lasers from ADS scenes and is proved to surpass SOTA performance.
- 4
- 4
- 4
- 4
- 4
- DC Coefficients Recovery from AC Coefficients in the JPEG Compression Scenario4
This paper proposes to highly enhance the JPEG compression by transmitting all AC coefficients and only four DC coefficients of one image and proposes two methods to rebuild the missing DC coefficients based on the remaining DCT coefficient relationships.
- 4
- Character as a Latent Variable in Large Language Models: A Mechanistic Account of Emergent Misalignment and Conditional Safety Failures3
Across multiple domains and model families, it is found that fine-tuning models on data exhibiting specific character-level dispositions induces substantially stronger and more transferable misalignment than incorrect-advice fine-tuning, while largely preserving general capabilities.
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- ObfusBFA: A Holistic Approach to Safeguarding DNNs From Different Types of Bit-Flip Attacks2
The key idea of ObfusBFA is to introduce random dummy operations during the model inference, which effectively transforms the delicate attacks into random bit flips, making it much harder for attackers to pinpoint and exploit vulnerable bits.
- 2
- 2
- 2
- Operationalizing AI in Future Networks: A Bird's Eye View from the System Perspective2
This article enumerate the key factors hindering the integration of AI/ML in real networks and review existing solutions to uncover the missing considerations, and highlights a promising direction, i.e., Machine Learning Operations (MLOps), that can close the gap.
- A Unified Hardware-based Threat Detector for AI Accelerators2
UniGuard, a novel unified and non-intrusive detection methodology to safeguard FPGA-based AI accelerators, is designed to harness power side-channel information generated during model inference to spot any anomaly.
- 2
- Survive at All Costs: Exploring LLM's Risky Behaviors under Survival Pressure1
A significant prevalence of SURVIVE-AT-ALL-COSTS misbehaviors in current models is revealed, the tangible real-world impact it may have, and insights for potential detection and mitigation strategies are provided.
- Picky LLMs and Unreliable RMs: An Empirical Study on Safety Alignment after Instruction Tuning1
This study systematically examines the factors contributing to safety alignment degradation in benign fine-tuning scenarios and evaluates the reliability of state-of-the-art reward models (RMs), which are often used to guide alignment processes.
- 1
- Can Large Language Models Automate the Refinement of Cellular Network Specifications?1
This paper investigates the application of Large Language Models for automated cellular network specification refinement and identifies Change Requests, which record specification revisions, as a key source of domain-specific data and formulate specification refinement as three complementary sub-tasks.
- 1
- 1
- 1
- VISO: Accelerating In-Orbit Object Detection with Language-Guided Mask Learning and Sparse Inference1
- Fingerprinting Image-to-Image Generative Adversarial Networks1
A novel finger-printing scheme for the Intellectual Property (IP) protection of image-to-image GANs based on a trusted third party and proves that these methods can satisfy different security requirements necessary for IP protection.
- 1
- Rethinking Adversarial Training with Neural Tangent Kernel1
An in-depth investigation of AT process and properties with NTK, such as NTK evolution, and the impact of data normalization on AT and the importance of unbiased estimators in batch normalization layers is disclosed.
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- –
- LeakDojo: Decoding the Leakage Threats of RAG Systems–
This study reveals that query generation and adversarial instructions contribute independently to leakage, with overall leakage well approximated by their product, and stronger instruction-following capability correlates with higher leakage risk.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Visible Yet Unreadable: A Systematic Blind Spot of Vision Language Models Across Writing Systems–
This paper constructs two psychophysics inspired benchmarks across distinct writing systems, Chinese logographs and English alphabetic words, by splicing, recombining, and overlaying glyphs to yield visible but unreadablestimuli for models while remaining legible to humans.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.