Low-light image enhancement aims to improve contrast and detail representation under insufficient illumination. However, existing methods primarily rely on deepening or widening convolutional layers, while neglecting image prior information, often resulting in detail loss and visual distortion. Moreover, convolutional neural networks (CNNs) struggle to capture long-range dependencies. To address these limitations, we propose a Gradient-Aware Transformer (GAformer), which integrates gradient-aware convolutions with Transformer-based global modelling. By leveraging gradient priors for enhanced local structural representation and exploiting the global interaction capability of Transformers, GAformer achieves more comprehensive and stable enhancement. Specifically, Gradient-Aware Kernels (GAK) are introduced to optimise edge feature extraction, followed by an Illumination Map-Guided Attention (IGA) mechanism that selectively enhances low-illumination regions. Furthermore, a Frequency-Modulated Calibration (FMC) module facilitates interaction between low- and high-frequency components for progressive guided recovery. Experimental results on multiple benchmark datasets demonstrate that GAformer outperforms state-of-the-art methods in both quantitative evaluation and visual quality.

Achieving counterfactual fairness is a critical objective in advancing fairness research within machine learning. Studies have shown that machine learning models often inherit biases from their training data, leading to unfair decision-making. Fair data generation methods aim to mitigate these biases, ensuring that predictors trained on such data uphold fairness. However, in the context of counterfactual fairness, existing methods for generating fair data are often limited in their applicability and lead to significant performance losses in downstream predictors. To address these issues, this paper proposes a new algorithm for generating counterfactually fair data, allowing predictors trained on this generated data to adhere to counterfactual fairness. We propose a new metric, Intervention-Driven Correlation (IDC), to evaluate the counterfactual fairness of generative models. IDC assesses fairness by applying random interventions to samples and measuring the statistical correlation between the degree of intervention and the outcome of interest. This metric is applicable to both discrete and continuous sensitive attributes and labels. Furthermore, our studies reveal a critical insight: counterfactually fair data does not always guarantee counterfactually fair predictors when deployed in real-world scenarios. We identify the root causes of this issue and propose a robust solution. To bridge this gap, we propose the IDC-Reduction method, which ensures the fairness of downstream predictors by generating counterfactually fair data. Experimentally, our method outperforms existing approaches and achieves counterfactual fairness regardless of the type of downstream predictors.

Transformer models are increasingly deployed in critical applications, raising concerns about their security and robustness. Like convolutional neural networks (CNNs), transformers are vulnerable to adversarial attacks, including Universal Adversarial Perturbations (UAP), which mislead models with a single perturbation across multiple inputs. However, existing UAP methods are primarily designed for CNNs and struggle with Transformer architectures due to their self-attention mechanisms, leaving potential vulnerabilities underexplored. To address these limitations, we propose Attention-Shift Universal Adversarial Perturbation (AS-UAP), a novel UAP method specifically designed for transformer-based models. AS-UAP generates perturbations that mislead the model’s attention toward irrelevant regions by leveraging attention patterns, without requiring label information. It introduces a new way to identify distracted attention heads based on their contribution to the [CLS] token during the forward pass, rather than relying on gradients. By leveraging the redundancy and operational patterns of attention heads, AS-UAP effectively misdirects the model’s attention, resulting in misclassification. Experimental results demonstrate that AS-UAP improves attack effectiveness, achieving a 11.23% higher fooling rate than the best existing UAP method. This study systematically reveals universal adversarial vulnerabilities in Transformer architectures, highlighting exploitable security gaps. These insights contribute to the advancement of AI security, ultimately improving the dependability of AI systems in mission-critical applications.

Federated learning (FL) is a distributed machine learning framework that enables multiple participants to train models without directly sharing local data. However, sensitive information about participants may still be leaked through their gradients. Furthermore, centralized servers used for aggregating these gradients can be vulnerable to compromise, leading to privacy violations or other malicious attacks. Therefore, it is essential to verify the integrity of the aggregation. In this work, we focus on designing communication efficient and fast verifiable aggregations for FL. We propose VSAFL, a verifiable secure aggregation (SecAgg) protocol specifically designed for cross-device FL. VSAFL achieves computation and communication cost of$O(\log ^{2} n + l \log n)$and$O(\log n + l)$, respectively, for SecAgg and verification for each user in each epoch, where n represents the number of clients and l denotes the dimension of the gradient vector. By employing a lightweight cryptographic primitive pseudorandom generator, VSAFL enables central servers and clients to prove and verify the correctness of model aggregations, significantly reducing verification costs. Our polynomial logarithmic overhead is particularly advantageous for clients with limited resources and high-dimensional gradients. Additionally, the proposed protocol is to be fully robust to clients dropping at any point. Through experimental evaluation, we demonstrate that VSAFL outperforms prior work in terms of verification speed by orders of magnitude.

Incentive-based applications enable users to obtain rewards after they complete a task, but how to balance the privacy and accountability is one of the most serious concerns currently. Publicly accountable anonymous authentication provides an excellent way verifying a user’s identity in a privacy-preserving way while ensuring public accountability in case of dispute. Although different kinds of these schemes have been proposed, they all assume that there is a single centralized certificate authority issuing a certificate to users, and are not suitable for an actual scenario which always involves multiple authorities. Therefore, a new primitive called multi-authority linkable and traceable anonymous authentication is proposed to address this issue, enabling privacy protection while holding public accountability under a multi-authority setting. We formally define a security model for this new notion and simultaneously design a generic construction while giving the security proof. Additionally, we implement the proposed scheme to show its efficiency.

As a useful cryptographic primitive, searchable symmetric encryption (SSE) has been intensively studied to achieve the secure and efficient retrieval of encrypted data. In order to process update operations, dynamic SSE schemes have been proposed. But recently, file-injection attack has threatened the security of traditional dynamic SSE protocols. Therefore, designing dynamic SSE schemes with forward privacy becomes a new demand to resist the above attack. Meanwhile, multi-client setting is another requirement in SSE techniques where multiple clients can be delegated and have access to the database. However, most of previous forward private schemes were constructed for single-client setting and cannot directly extended to multi-client environment efficiently. To solve the problem, we propose a forward private SSE scheme with support for multi-client in cloud computing. The proposed scheme is based on XOR-homomorphic function and involves two new data structures as private link and public search tree. Security proof demonstrates the proposed scheme can meet the desired secure features. We then conduct experimental evaluation of the proposed scheme and make comparison with related schemes. The result shows that the proposed scheme tends to have high efficiency.

While many test input generation techniques have been proposed to improve the code coverage of dynamic analysis, they are still inefficient in triggering hidden malicious behaviors protected by anti-analysis techniques. In this work, we design and implement Droid-AntiRM, a new approach seeking to tame anti-analysis automatically and improve automated dynamic analysis. Our approach leverages three key observations: 1) Logic-bomb based anti-analysis techniques control the execution of certain malicious behaviors; 2) Anti-analysis techniques are normally implemented through condition statements; 3) Anti-analysis techniques normally have no dependence on program inputs. Based on these observations, Droid-AntiRM uses various techniques to detect anti-analysis in malware samples, and rewrite the condition statements in anti-analysis cases through bytecode instrumentation, thus forcing the hidden behavior to be executed at runtime. Through a study of 3187 malware samples, we find that 32.50% of them employ various anti-analysis techniques. Our experiments demonstrate that Droid-AntiRM can identify anti-analysis instances from 30 malware samples with a true positive rate of 89.15% and zero false negative. By taming the identified anti-analysis, Droid-AntiRM can greatly improve the automated dynamic analysis, successfully triggering 44 additional hidden malicious behaviors from the 30 samples. Further performance evaluation shows that Droid-AntiRM has good efficiency to perform large-scale analysis.

Certificateless signature is an interesting cryptographic primitive which does not suffer from the inherent key escrow problem of identity‐based cryptography and the costly certificate management problem of traditional public key cryptography. Since security proofs in the random oracle model can only be viewed as heuristic arguments and cannot ensure the security in the real implementation, certificateless signature schemes with security proofs in the standard model (i.e. without random oracles) is more desirable. Some attempts have been devoted to propose certificateless signature schemes in the standard model, whereas all of these schemes are later shown to be either insecure or flawed in the security proofs. Recently, a new certificateless signature scheme in the standard model has been proposed. However, in this study the authors show that this scheme cannot resist the key replacement attack, and hence it is not existentially unforgeable.