arXiv subjects

Hua Ma

Publications and source records attributed to Hua Ma.

At least 19 recordsLinked to original sources

arXiv2026

iTIMO: An LLM-empowered Synthesis Dataset for Travel Itinerary Modification

Addressing itinerary modification is crucial for enhancing the travel experience as it is a frequent requirement during traveling. However, existing research mainly focuses on fixed itinerary planning, leaving modification underexplored due to the scarcity of need-to-modify itinerary data. To bridge this gap, we formally define the itinerary modification task and propose a general pipeline to construct the corresponding dataset, namely iTIMO. This pipeline frames the generation of need-to-modify itinerary data as an intent-driven perturbation task. It instructs large language models to perturb real-world itineraries using three operations: REPLACE, ADD, and DELETE. Each perturbation is grounded in three intents: disruptions of popularity, spatial distance, and category diversity. Furthermore, hybrid evaluation metrics are introduced to ensure perturbation effectiveness. We conduct comprehensive benchmarking on iTIMO to analyze the capabilities and limitations of state-of-the-art LLMs. Overall, iTIMO provides a comprehensive testbed for the modification task, and empowers the evolution of traditional travel recommender systems into adaptive frameworks capable of handling dynamic travel needs. Dataset, code and supplementary materials are available at https://github.com/zelo2/iTIMO.

cs.IR↗

arXiv2025

Keep the Lights On, Keep the Lengths in Check: Plug-In Adversarial Detection for Time-Series LLMs in Energy Forecasting

Accurate time-series forecasting is increasingly critical for planning and operations in low-carbon power systems. Emerging time-series large language models (TS-LLMs) now deliver this capability at scale, requiring no task-specific retraining, and are quickly becoming essential components within the Internet-of-Energy (IoE) ecosystem. However, their real-world deployment is complicated by a critical vulnerability: adversarial examples (AEs). Detecting these AEs is challenging because (i) adversarial perturbations are optimized across the entire input sequence and exploit global temporal dependencies, which renders local detection methods ineffective, and (ii) unlike traditional forecasting models with fixed input dimensions, TS-LLMs accept sequences of variable length, increasing variability that complicates detection. To address these challenges, we propose a plug-in detection framework that capitalizes on the TS-LLM's own variable-length input capability. Our method uses sampling-induced divergence as a detection signal. Given an input sequence, we generate multiple shortened variants and detect AEs by measuring the consistency of their forecasts: Benign sequences tend to produce stable predictions under sampling, whereas adversarial sequences show low forecast similarity, because perturbations optimized for a full-length sequence do not transfer reliably to shorter, differently-structured subsamples. We evaluate our approach on three representative TS-LLMs (TimeGPT, TimesFM, and TimeLLM) across three energy datasets: ETTh2 (Electricity Transformer Temperature), NI (Hourly Energy Consumption), and Consumption (Hourly Electricity Consumption and Production). Empirical results confirm strong and robust detection performance across both black-box and white-box attack scenarios, highlighting its practicality as a reliable safeguard for TS-LLM forecasting in real-world energy systems.

cs.CR↗

arXiv2025

Supervised Fine Tuning of Large Language Models for Domain Specific Knowledge Graph Construction:A Case Study on Hunan's Historical Celebrities

Large language models and knowledge graphs offer strong potential for advancing research on historical culture by supporting the extraction, analysis, and interpretation of cultural heritage. Using Hunan's modern historical celebrities shaped by Huxiang culture as a case study, pre-trained large models can help researchers efficiently extract key information, including biographical attributes, life events, and social relationships, from textual sources and construct structured knowledge graphs. However, systematic data resources for Hunan's historical celebrities remain limited, and general-purpose models often underperform in domain knowledge extraction and structured output generation in such low-resource settings. To address these issues, this study proposes a supervised fine-tuning approach for enhancing domain-specific information extraction. First, we design a fine-grained, schema-guided instruction template tailored to the Hunan historical celebrities domain and build an instruction-tuning dataset to mitigate the lack of domain-specific training corpora. Second, we apply parameter-efficient instruction fine-tuning to four publicly available large language models - Qwen2.5-7B, Qwen3-8B, DeepSeek-R1-Distill-Qwen-7B, and Llama-3.1-8B-Instruct - and develop evaluation criteria for assessing their extraction performance. Experimental results show that all models exhibit substantial performance gains after fine-tuning. Among them, Qwen3-8B achieves the strongest results, reaching a score of 89.3866 with 100 samples and 50 training iterations. This study provides new insights into fine-tuning vertical large language models for regional historical and cultural domains and highlights their potential for cost-effective applications in cultural heritage knowledge extraction and knowledge graph construction.

cs.CL↗

arXiv2025

Kill Two Birds with One Stone! Trajectory enabled Unified Online Detection of Adversarial Examples and Backdoor Attacks

The proposed UniGuard is the first unified online detection framework capable of simultaneously addressing adversarial examples and backdoor attacks. UniGuard builds upon two key insights: first, both AE and backdoor attacks have to compromise the inference phase, making it possible to tackle them simultaneously during run-time via online detection. Second, an adversarial input, whether a perturbed sample in AE attacks or a trigger-carrying sample in backdoor attacks, exhibits distinctive trajectory signatures from a benign sample as it propagates through the layers of a DL model in forward inference. The propagation trajectory of the adversarial sample must deviate from that of its benign counterpart; otherwise, the adversarial objective cannot be fulfilled. Detecting these trajectory signatures is inherently challenging due to their subtlety; UniGuard overcomes this by treating the propagation trajectory as a time-series signal, leveraging LSTM and spectrum transformation to amplify differences between adversarial and benign trajectories that are subtle in the time domain. UniGuard exceptional efficiency and effectiveness have been extensively validated across various modalities (image, text, and audio) and tasks (classification and regression), ranging from diverse model architectures against a wide range of AE attacks and backdoor attacks, including challenging partial backdoors and dynamic triggers. When compared to SOTA methods, including ContraNet (NDSS 22) specific for AE detection and TED (IEEE SP 24) specific for backdoor detection, UniGuard consistently demonstrates superior performance, even when matched against each method's strengths in addressing their respective threats-each SOTA fails to parts of attack strategies while UniGuard succeeds for all.

cs.CR↗

arXiv2025

From Pixels to Trajectory: Universal Adversarial Example Detection via Temporal Imprints

For the first time, we unveil discernible temporal (or historical) trajectory imprints resulting from adversarial example (AE) attacks. Standing in contrast to existing studies all focusing on spatial (or static) imprints within the targeted underlying victim models, we present a fresh temporal paradigm for understanding these attacks. Of paramount discovery is that these imprints are encapsulated within a single loss metric, spanning universally across diverse tasks such as classification and regression, and modalities including image, text, and audio. Recognizing the distinct nature of loss between adversarial and clean examples, we exploit this temporal imprint for AE detection by proposing TRAIT (TRaceable Adversarial temporal trajectory ImprinTs). TRAIT operates under minimal assumptions without prior knowledge of attacks, thereby framing the detection challenge as a one-class classification problem. However, detecting AEs is still challenged by significant overlaps between the constructed synthetic losses of adversarial and clean examples due to the absence of ground truth for incoming inputs. TRAIT addresses this challenge by converting the synthetic loss into a spectrum signature, using the technique of Fast Fourier Transform to highlight the discrepancies, drawing inspiration from the temporal nature of the imprints, analogous to time-series signals. Across 12 AE attacks including SMACK (USENIX Sec'2023), TRAIT demonstrates consistent outstanding performance across comprehensively evaluated modalities, tasks, datasets, and model architectures. In all scenarios, TRAIT achieves an AE detection accuracy exceeding 97%, often around 99%, while maintaining a false rejection rate of 1%. TRAIT remains effective under the formulated strong adaptive attacks.

cs.CR↗

arXiv2025

Comprehensive Evaluation of Cloaking Backdoor Attacks on Object Detector in Real-World

The exploration of backdoor vulnerabilities in object detectors, particularly in real-world scenarios, remains limited. A significant challenge lies in the absence of a natural physical backdoor dataset, and constructing such a dataset is both time- and labor-intensive. In this work, we address this gap by creating a large-scale dataset comprising approximately 11,800 images/frames with annotations featuring natural objects (e.g., T-shirts and hats) as triggers to incur cloaking adversarial effects in diverse real-world scenarios. This dataset is tailored for the study of physical backdoors in object detectors. Leveraging this dataset, we conduct a comprehensive evaluation of an insidious cloaking backdoor effect against object detectors, wherein the bounding box around a person vanishes when the individual is near a natural object (e.g., a commonly available T-shirt) in front of the detector. Our evaluations encompass three prevalent attack surfaces: data outsourcing, model outsourcing, and the use of pretrained models. The cloaking effect is successfully implanted in object detectors across all three attack surfaces. We extensively evaluate four popular object detection algorithms (anchor-based Yolo-V3, Yolo-V4, Faster R-CNN, and anchor-free CenterNet) using 19 videos (totaling approximately 11,800 frames) in real-world scenarios. Our results demonstrate that the backdoor attack exhibits remarkable robustness against various factors, including movement, distance, angle, non-rigid deformation, and lighting. In data and model outsourcing scenarios, the attack success rate (ASR) in most videos reaches 100% or near it, while the clean data accuracy of the backdoored model remains indistinguishable from that of the clean model, making it impossible to detect backdoor behavior through a validation set.

cs.CR↗

arXiv2024

qMRI Diffuser: Quantitative T1 Mapping of the Brain using a Denoising Diffusion Probabilistic Model

Quantitative MRI (qMRI) offers significant advantages over weighted images by providing objective parameters related to tissue properties. Deep learning-based methods have demonstrated effectiveness in estimating quantitative maps from series of weighted images. In this study, we present qMRI Diffuser, a novel approach to qMRI utilising deep generative models. Specifically, we implemented denoising diffusion probabilistic models (DDPM) for T1 quantification in the brain, framing the estimation of quantitative maps as a conditional generation task. The proposed method is compared with the residual neural network (ResNet) and the recurrent inference machine (RIM) on both phantom and in vivo data. The results indicate that our method achieves improved accuracy and precision in parameter estimation, along with superior visual performance. Moreover, our method inherently incorporates stochasticity, enabling straightforward quantification of uncertainty. Hence, the proposed method holds significant promise for quantitative MR mapping.

cs.CV↗

arXiv2023

Watch Out! Simple Horizontal Class Backdoor Can Trivially Evade Defense

All current backdoor attacks on deep learning (DL) models fall under the category of a vertical class backdoor (VCB) -- class-dependent. In VCB attacks, any sample from a class activates the implanted backdoor when the secret trigger is present. Existing defense strategies overwhelmingly focus on countering VCB attacks, especially those that are source-class-agnostic. This narrow focus neglects the potential threat of other simpler yet general backdoor types, leading to false security implications. This study introduces a new, simple, and general type of backdoor attack coined as the horizontal class backdoor (HCB) that trivially breaches the class dependence characteristic of the VCB, bringing a fresh perspective to the community. HCB is now activated when the trigger is presented together with an innocuous feature, regardless of class. For example, the facial recognition model misclassifies a person who wears sunglasses with a smiling innocuous feature into the targeted person, such as an administrator, regardless of which person. The key is that these innocuous features are horizontally shared among classes but are only exhibited by partial samples per class. Extensive experiments on attacking performance across various tasks, including MNIST, facial recognition, traffic sign recognition, object detection, and medical diagnosis, confirm the high efficiency and effectiveness of the HCB. We rigorously evaluated the evasiveness of the HCB against a series of eleven representative countermeasures, including Fine-Pruning (RAID 18'), STRIP (ACSAC 19'), Neural Cleanse (Oakland 19'), ABS (CCS 19'), Februus (ACSAC 20'), NAD (ICLR 21'), MNTD (Oakland 21'), SCAn (USENIX SEC 21'), MOTH (Oakland 22'), Beatrix (NDSS 23'), and MM-BD (Oakland 24'). None of these countermeasures prove robustness, even when employing a simplistic trigger, such as a small and static white-square patch.

cs.CR↗

arXiv2023

DeepTheft: Stealing DNN Model Architectures through Power Side Channel

Deep Neural Network (DNN) models are often deployed in resource-sharing clouds as Machine Learning as a Service (MLaaS) to provide inference services.To steal model architectures that are of valuable intellectual properties, a class of attacks has been proposed via different side-channel leakage, posing a serious security challenge to MLaaS. Also targeting MLaaS, we propose a new end-to-end attack, DeepTheft, to accurately recover complex DNN model architectures on general processors via the RAPL-based power side channel. However, an attacker can acquire only a low sampling rate (1 KHz) of the time-series energy traces from the RAPL interface, rendering existing techniques ineffective in stealing large and deep DNN models. To this end, we design a novel and generic learning-based framework consisting of a set of meta-models, based on which DeepTheft is demonstrated to have high accuracy in recovering a large number (thousands) of models architectures from different model families including the deepest ResNet152. Particularly, DeepTheft has achieved a Levenshtein Distance Accuracy of 99.75% in recovering network structures, and a weighted average F1 score of 99.60% in recovering diverse layer-wise hyperparameters. Besides, our proposed learning framework is general to other time-series side-channel signals. To validate its generalization, another existing side channel is exploited, i.e., CPU frequency. Different from RAPL, CPU frequency is accessible to unprivileged users in bare-metal OSes. By using our generic learning framework trained against CPU frequency traces, DeepTheft has shown similarly high attack performance in stealing model architectures.

cs.CR↗

arXiv2023

One-to-Multiple Clean-Label Image Camouflage (OmClic) based Backdoor Attack on Deep Learning

Image camouflage has been utilized to create clean-label poisoned images for implanting backdoor into a DL model. But there exists a crucial limitation that one attack/poisoned image can only fit a single input size of the DL model, which greatly increases its attack budget when attacking multiple commonly adopted input sizes of DL models. This work proposes to constructively craft an attack image through camouflaging but can fit multiple DL models' input sizes simultaneously, namely OmClic. Thus, through OmClic, we are able to always implant a backdoor regardless of which common input size is chosen by the user to train the DL model given the same attack budget (i.e., a fraction of the poisoning rate). With our camouflaging algorithm formulated as a multi-objective optimization, M=5 input sizes can be concurrently targeted with one attack image, which artifact is retained to be almost visually imperceptible at the same time. Extensive evaluations validate the proposed OmClic can reliably succeed in various settings using diverse types of images. Further experiments on OmClic based backdoor insertion to DL models show that high backdoor performances (i.e., attack success rate and clean data accuracy) are achievable no matter which common input size is randomly chosen by the user to train the model. So that the OmClic based backdoor attack budget is reduced by M$\times$ compared to the state-of-the-art camouflage based backdoor attack as a baseline. Significantly, the same set of OmClic based poisonous attack images is transferable to different model architectures for backdoor implant.

cs.CR↗

arXiv2023

TransCAB: Transferable Clean-Annotation Backdoor to Object Detection with Natural Trigger in Real-World

Object detection is the foundation of various critical computer-vision tasks such as segmentation, object tracking, and event detection. To train an object detector with satisfactory accuracy, a large amount of data is required. However, due to the intensive workforce involved with annotating large datasets, such a data curation task is often outsourced to a third party or relied on volunteers. This work reveals severe vulnerabilities of such data curation pipeline. We propose MACAB that crafts clean-annotated images to stealthily implant the backdoor into the object detectors trained on them even when the data curator can manually audit the images. We observe that the backdoor effect of both misclassification and the cloaking are robustly achieved in the wild when the backdoor is activated with inconspicuously natural physical triggers. Backdooring non-classification object detection with clean-annotation is challenging compared to backdooring existing image classification tasks with clean-label, owing to the complexity of having multiple objects within each frame, including victim and non-victim objects. The efficacy of the MACAB is ensured by constructively i abusing the image-scaling function used by the deep learning framework, ii incorporating the proposed adversarial clean image replica technique, and iii combining poison data selection criteria given constrained attacking budget. Extensive experiments demonstrate that MACAB exhibits more than 90% attack success rate under various real-world scenes. This includes both cloaking and misclassification backdoor effect even restricted with a small attack budget. The poisoned samples cannot be effectively identified by state-of-the-art detection techniques.The comprehensive video demo is at https://youtu.be/MA7L_LpXkp4, which is based on a poison rate of 0.14% for YOLOv4 cloaking backdoor and Faster R-CNN misclassification backdoor.

cs.CV↗

arXiv2023

CIT-EmotionNet: CNN Interactive Transformer Network for EEG Emotion Recognition

Emotion recognition using Electroencephalogram (EEG) signals has emerged as a significant research challenge in affective computing and intelligent interaction. However, effectively combining global and local features of EEG signals to improve performance in emotion recognition is still a difficult task. In this study, we propose a novel CNN Interactive Transformer Network for EEG Emotion Recognition, known as CIT-EmotionNet, which efficiently integrates global and local features of EEG signals. Initially, we convert raw EEG signals into spatial-frequency representations, which serve as inputs. Then, we integrate Convolutional Neural Network (CNN) and Transformer within a single framework in a parallel manner. Finally, we design a CNN interactive Transformer module, which facilitates the interaction and fusion of local and global features, thereby enhancing the model's ability to extract both types of features from EEG spatial-frequency representations. The proposed CIT-EmotionNet outperforms state-of-the-art methods, achieving an average recognition accuracy of 98.57\% and 92.09\% on two publicly available datasets, SEED and SEED-IV, respectively.

eess.SP↗

arXiv2023

Quantization Backdoors to Deep Learning Commercial Frameworks

Currently, there is a burgeoning demand for deploying deep learning (DL) models on ubiquitous edge Internet of Things (IoT) devices attributed to their low latency and high privacy preservation. However, DL models are often large in size and require large-scale computation, which prevents them from being placed directly onto IoT devices, where resources are constrained and 32-bit floating-point (float-32) operations are unavailable. Commercial framework (i.e., a set of toolkits) empowered model quantization is a pragmatic solution that enables DL deployment on mobile devices and embedded systems by effortlessly post-quantizing a large high-precision model (e.g., float-32) into a small low-precision model (e.g., int-8) while retaining the model inference accuracy. However, their usability might be threatened by security vulnerabilities. This work reveals that the standard quantization toolkits can be abused to activate a backdoor. We demonstrate that a full-precision backdoored model which does not have any backdoor effect in the presence of a trigger -- as the backdoor is dormant -- can be activated by the default i) TensorFlow-Lite (TFLite) quantization, the only product-ready quantization framework to date, and ii) the beta released PyTorch Mobile framework. When each of the float-32 models is converted into an int-8 format model through the standard TFLite or Pytorch Mobile framework's post-training quantization, the backdoor is activated in the quantized model, which shows a stable attack success rate close to 100% upon inputs with the trigger, while it behaves normally upon non-trigger inputs. This work highlights that a stealthy security threat occurs when an end user utilizes the on-device post-training model quantization frameworks, informing security researchers of cross-platform overhaul of DL models post quantization even if these models pass front-end backdoor inspections.

cs.CR↗

arXiv2023

Vertical Federated Learning: Taxonomies, Threats, and Prospects

Federated learning (FL) is the most popular distributed machine learning technique. FL allows machine-learning models to be trained without acquiring raw data to a single point for processing. Instead, local models are trained with local data; the models are then shared and combined. This approach preserves data privacy as locally trained models are shared instead of the raw data themselves. Broadly, FL can be divided into horizontal federated learning (HFL) and vertical federated learning (VFL). For the former, different parties hold different samples over the same set of features; for the latter, different parties hold different feature data belonging to the same set of samples. In a number of practical scenarios, VFL is more relevant than HFL as different companies (e.g., bank and retailer) hold different features (e.g., credit history and shopping history) for the same set of customers. Although VFL is an emerging area of research, it is not well-established compared to HFL. Besides, VFL-related studies are dispersed, and their connections are not intuitive. Thus, this survey aims to bring these VFL-related studies to one place. Firstly, we classify existing VFL structures and algorithms. Secondly, we present the threats from security and privacy perspectives to VFL. Thirdly, for the benefit of future researchers, we discussed the challenges and prospects of VFL in detail.

cs.LG↗

arXiv2022

RBNN: Memory-Efficient Reconfigurable Deep Binary Neural Network with IP Protection for Internet of Things

Though deep neural network models exhibit outstanding performance for various applications, their large model size and extensive floating-point operations render deployment on mobile computing platforms a major challenge, and, in particular, on Internet of Things devices. One appealing solution is model quantization that reduces the model size and uses integer operations commonly supported by microcontrollers . To this end, a 1-bit quantized DNN model or deep binary neural network maximizes the memory efficiency, where each parameter in a BNN model has only 1-bit. In this paper, we propose a reconfigurable BNN (RBNN) to further amplify the memory efficiency for resource-constrained IoT devices. Generally, the RBNN can be reconfigured on demand to achieve any one of M (M>1) distinct tasks with the same parameter set, thus only a single task determines the memory requirements. In other words, the memory utilization is improved by times M. Our extensive experiments corroborate that up to seven commonly used tasks can co-exist (the value of M can be larger). These tasks with a varying number of classes have no or negligible accuracy drop-off on three binarized popular DNN architectures including VGG, ResNet, and ReActNet. The tasks span across different domains, e.g., computer vision and audio domains validated herein, with the prerequisite that the model architecture can serve those cross-domain tasks. To protect the intellectual property of an RBNN model, the reconfiguration can be controlled by both a user key and a device-unique root key generated by the intrinsic hardware fingerprint. By doing so, an RBNN model can only be used per paid user per authorized device, thus benefiting both the user and the model provider.

cs.CR↗

arXiv2022

MUD-PQFed: Towards Malicious User Detection in Privacy-Preserving Quantized Federated Learning

Federated Learning (FL), a distributed machine learning paradigm, has been adapted to mitigate privacy concerns for customers. Despite their appeal, there are various inference attacks that can exploit shared-plaintext model updates to embed traces of customer private information, leading to serious privacy concerns. To alleviate this privacy issue, cryptographic techniques such as Secure Multi-Party Computation and Homomorphic Encryption have been used for privacy-preserving FL. However, such security issues in privacy-preserving FL are poorly elucidated and underexplored. This work is the first attempt to elucidate the triviality of performing model corruption attacks on privacy-preserving FL based on lightweight secret sharing. We consider scenarios in which model updates are quantized to reduce communication overhead in this case, where an adversary can simply provide local parameters outside the legal range to corrupt the model. We then propose the MUD-PQFed protocol, which can precisely detect malicious clients performing attacks and enforce fair penalties. By removing the contributions of detected malicious clients, the global model utility is preserved to be comparable to the baseline global model without the attack. Extensive experiments validate effectiveness in maintaining baseline accuracy and detecting malicious clients in a fine-grained manner

cs.CR↗

arXiv2022

Dangerous Cloaking: Natural Trigger based Backdoor Attacks on Object Detectors in the Physical World

Deep learning models have been shown to be vulnerable to recent backdoor attacks. A backdoored model behaves normally for inputs containing no attacker-secretly-chosen trigger and maliciously for inputs with the trigger. To date, backdoor attacks and countermeasures mainly focus on image classification tasks. And most of them are implemented in the digital world with digital triggers. Besides the classification tasks, object detection systems are also considered as one of the basic foundations of computer vision tasks. However, there is no investigation and understanding of the backdoor vulnerability of the object detector, even in the digital world with digital triggers. For the first time, this work demonstrates that existing object detectors are inherently susceptible to physical backdoor attacks. We use a natural T-shirt bought from a market as a trigger to enable the cloaking effect--the person bounding-box disappears in front of the object detector. We show that such a backdoor can be implanted from two exploitable attack scenarios into the object detector, which is outsourced or fine-tuned through a pretrained model. We have extensively evaluated three popular object detection algorithms: anchor-based Yolo-V3, Yolo-V4, and anchor-free CenterNet. Building upon 19 videos shot in real-world scenes, we confirm that the backdoor attack is robust against various factors: movement, distance, angle, non-rigid deformation, and lighting. Specifically, the attack success rate (ASR) in most videos is 100% or close to it, while the clean data accuracy of the backdoored model is the same as its clean counterpart. The latter implies that it is infeasible to detect the backdoor behavior merely through a validation set. The averaged ASR still remains sufficiently high to be 78% in the transfer learning attack scenarios evaluated on CenterNet. See the demo video on https://youtu.be/Q3HOF4OobbY.

cs.CV↗

arXiv2022

Towards A Critical Evaluation of Robustness for Deep Learning Backdoor Countermeasures

Since Deep Learning (DL) backdoor attacks have been revealed as one of the most insidious adversarial attacks, a number of countermeasures have been developed with certain assumptions defined in their respective threat models. However, the robustness of these countermeasures is inadvertently ignored, which can introduce severe consequences, e.g., a countermeasure can be misused and result in a false implication of backdoor detection. For the first time, we critically examine the robustness of existing backdoor countermeasures with an initial focus on three influential model-inspection ones that are Neural Cleanse (S&P'19), ABS (CCS'19), and MNTD (S&P'21). Although the three countermeasures claim that they work well under their respective threat models, they have inherent unexplored non-robust cases depending on factors such as given tasks, model architectures, datasets, and defense hyper-parameter, which are \textit{not even rooted from delicate adaptive attacks}. We demonstrate how to trivially bypass them aligned with their respective threat models by simply varying aforementioned factors. Particularly, for each defense, formal proofs or empirical studies are used to reveal its two non-robust cases where it is not as robust as it claims or expects, especially the recent MNTD. This work highlights the necessity of thoroughly evaluating the robustness of backdoor countermeasures to avoid their misleading security implications in unknown non-robust cases.

cs.CR↗