Object detection models are vulnerable to backdoor attacks, where attackers poison a small subset of training samples by embedding a predefined trigger to manipulate prediction. Detecting poisoned samples (i.e., those containing triggers) at test time can prevent backdoor activation. However, unlike image classification tasks, the unique characteristics of object detection—particularly its output of numerous objects—pose fresh challenges for backdoor detection. The complex attack effects (e.g., "ghost" object emergence or "vanishing" object) further render current defenses fundamentally inadequate. To this end, we design TRAnsformation Consistency Evaluation (TRACE), a brand-new method for detecting poisoned samples at test time in object detection. Our journey begins with two intriguing observations: 1) poisoned samples exhibit significantly more consistent detection results than clean ones across varied backgrounds. 2) clean samples show higher detection consistency when introduced to different focal information. Based on these phenomena, Trace applies foreground and background transformations to each test sample, then assesses transformation consistency by calculating the variance in objects confidences. Trace achieves black-box, universal backdoor detection, with extensive experiments showing a 30% improvement in AUROC over state-of-the-art defenses and resistance to adaptive attacks.

Deep neural networks (DNNs) are vulnerable to backdoor attacks, where adversaries manipulate the training data to implant hidden backdoors, causing misclassifications when inputs are modified with adversary-specified triggers. To ensure the reliability of DNNs, it is crucial to repair backdoored models, ensuring all samples, including poisoned ones, are correctly predicted. Existing backdoor repair methods rely on recovering a universal trigger, which degrades the model performance on clean data and fails against sophisticated attacks. In this paper, we propose FixGuard, a novel backdoor repair method based on a min-min optimization formulation. FixGuard consists of two main stages: the trigger recovery stage and the backdoor unlearning stage. In the trigger recovery stage, FixGuard recovers a semantic-aware trigger for each class. Compared to using a universal trigger, class-wise triggers can identify malicious patterns more precisely while avoiding excessively large triggers that degrade model performance on clean data. In the backdoor unlearning stage, FixGuard generates synthesized poisoned samples by patching clean images with the recovered triggers while preserving their correct labels. Subsequently, it trains backdoored models on these synthesized samples to sever the malicious connection between triggers and target labels. Experimental results demonstrate that FixGuard effectively mitigates backdoor threats while preserving model performance.

Deep neural networks (DNNs) are susceptible to backdoor attacks, where adversaries poison datasets with adversary-specified triggers to implant hidden backdoors, enabling malicious manipulation of model predictions. Dataset purification serves as a proactive defense by removing malicious training samples to prevent backdoor injection at its source. We first reveal that the current advanced purification methods rely on a latent assumption that the backdoor connections between triggers and target labels in backdoor attacks are simpler to learn than the benign features. We demonstrate that this assumption, however, does not always hold, especially in all-to-all (A2A) and untargeted (UT) attacks. As a result, purification methods that analyze the separation between the poisoned and benign samples in the input-output space or the final hidden layer space are less effective. We observe that this separability is not confined to a single layer but varies across different hidden layers. Motivated by this understanding, we propose FLARE, a universal purification method to counter various backdoor attacks. FLARE aggregates abnormal activations from all hidden layers to construct representations for clustering. To enhance separation, FLARE develops an adaptive subspace selection algorithm to isolate the optimal space for dividing an entire dataset into two clusters. FLARE assesses the stability of each cluster and identifies the cluster with higher stability as poisoned. Extensive evaluations on benchmark datasets demonstrate the effectiveness of FLARE against 22 representative backdoor attacks, including all-to-one (A2O), all-to-all (A2A), and untargeted (UT) attacks, and its robustness to adaptive attacks.

Deep neural networks (DNNs) are vulnerable to backdoor attacks, where adversaries can maliciously trigger model misclassifications by implanting a hidden backdoor during model training. This paper proposes a simple yet effective input-level backdoor detection (dubbed IBD-PSC) as a ‘firewall’ to filter out malicious testing images. Our method is motivated by an intriguing phenomenon, i.e., parameter-oriented scaling consistency (PSC), where the prediction confidences of poisoned samples are significantly more consistent than those of benign ones when amplifying model parameters. In particular, we provide theoretical analysis to safeguard the foundations of the PSC phenomenon. We also design an adaptive method to select BN layers to scale up for effective detection. Extensive experiments are conducted on benchmark datasets, verifying the effectiveness and efficiency of our IBD-PSC method and its resistance to adaptive attacks. Codes are available at https://github.com/THUYimingLi/BackdoorBox.

Deep neural networks (DNNs) are vulnerable to backdoor attacks, where a backdoored model behaves normally with clean inputs but exhibits attacker-specified behaviors upon the inputs containing triggers. Most previous backdoor attacks mainly focus on either the all-to-one or all-to-all paradigm, allowing attackers to manipulate an input to attack a single target class. Besides, the two paradigms rely on a single trigger for backdoor activation, rendering attacks ineffective if the trigger is destroyed. In light of the above, we propose a new M-to-N attack paradigm that allows an attacker to manipulate any input to attack N target classes, and each backdoor of the N target classes can be activated by any one of its M triggers. Our attack selects M clean images from each target class as triggers and leverages our proposed poisoned image generation framework to inject the triggers into clean images invisibly. By using triggers with the same distribution as clean training images, the targeted DNN models can generalize to the triggers during training, thereby enhancing the effectiveness of our attack on multiple target classes. Extensive experimental results demonstrate that our new backdoor attack is highly effective in attacking multiple target classes and robust against pre-processing operations and existing defenses.