3D visual grounding (3DVG) identifies objects in 3D scenes from language descriptions. Existing zero-shot approaches leverage 2D vision–language models (VLMs) by converting 3D spatial information (SI) into forms amenable to VLM processing, typically as composite inputs such as specified-view renderings or video sequences with overlaid object markers. However, this VLM ⊕ SI paradigm yields entangled visual representations that compel the VLM to process entire cluttered cues, making it hard to exploit spatial–semantic relationships effectively. In this work, we propose a new VLM ⊗ SI paradigm that externalizes the 3D SI into a form enabling the VLM to incrementally retrieve only what it needs during reasoning. We instantiate this paradigm with a novel View-on-Graph (VoG) method, which organizes the scene into a multi-modal, multi-layer scene graph and allows the VLM to operate as an active agent that selectively accesses necessary cues as it traverses the scene. This design offers two intrinsic advantages: (i) by structuring 3D context into a spatially and semantically coherent scene graph rather than confounding the VLM with densely entangled visual inputs, it lowers the VLM's reasoning difficulty; and (ii) by actively exploring and reasoning over the scene graph, it naturally produces transparent, step-by-step traces for interpretable 3DVG. Extensive experiments show that VoG achieves state-of-the-art zero-shot performance, establishing structured scene exploration as a promising strategy for advancing zero-shot 3DVG.

Automatic exploit generation (AEG) refers to the automatic discovery and exploitation of vulnerabilities against unknown targets. Traditional AEG often targets a single type of vulnerability and still relies on templates built from expert experience. To achieve intelligent exploit generation, we establish a comprehensive benchmark using Binary Exploitation (pwn) challenges in Capture the Flag (CTF) competitions and investigate the capabilities of Large Language Models (LLMs) in AEG based on the benchmark. To improve the performance of AEG, we propose PwnGPT, an LLM-based automatic exploit generation framework that automatically solves pwn challenges. The structural design of PwnGPT is divided into three main components: analysis, generation, and verification modules. With the help of a modular approach and structured problem inputs, PwnGPT can solve challenges that LLMs cannot directly solve. We evaluate PwnGPT on our benchmark and analyze the outputs of each module. Experimental results show that our framework is highly autonomous and capable of addressing various challenges. Compared to direct input LLMs, PwnGPT increases the completion rate of exploit on our benchmark from 26.3% to 57.9% with the OpenAI o1-preview model and from 21.1% to 36.8% with the GPT-4o model.

Industrial control systems (ICSs) are widely used in industry, and their security and stability are very important. Once the ICS is attacked, it may cause serious damage. Therefore, it is very important to detect anomalies in ICSs. ICS can monitor and manage physical devices remotely using communication networks. The existing anomaly detection approaches mainly focus on analyzing the security of network traffic or sensor data. However, the behaviors of different domains (e.g., network traffic and sensor physical status) of ICSs are correlated, so it is difficult to comprehensively identify anomalies by analyzing only a single domain. In this article, an anomaly detection approach based on cross-domain representation learning in ICSs is proposed, which can learn the joint features of multi-domain behaviors and detect anomalies within different domains. After constructing a cross-domain graph that can represent the behaviors of multiple domains in ICSs, our approach can learn the joint features of them by leveraging graph neural networks. Since anomalies behave differently in different domains, we leverage a multi-task learning approach to identify anomalies in different domains separately and perform joint training. The experimental results show that the performance of our approach is better than existing approaches for identifying anomalies in ICSs.

Sequence-based deep learning models are commonly used to detect anomalies in system logs to ensure the security of communication and information systems. However, recent research has shown that adversarial attack methods against these detection models reveal their vulnerabilities. Attackers can bypass these sequence-based classifiers by tampering with the sequence (e.g., adding or replacing sequence events). In this paper, we propose a novel Multi-Stage Defensive strategy for sequence-based log anomaly detection aimed at combating adversarial attacks. Systematically integrated, this strategy spans the entire detection process, from data embedding representation to anomaly classification, thereby forming a robust defensive approach that is strategically orchestrated to ensure comprehensive protection against a variety of adversarial attacks. Firstly, we compress the semantic feature space of sequences to enhance the anti-interference ability of attack operations on substitution sequences. Then, we propose a novel adaptive sparse multi-head attention mechanism to improve the Transformer model, allowing it to adaptively extract different key patterns in the sequence, thus eliminating irrelevant sequence events added in adversarial sequences. Our approach also integrates the learning of temporal patterns, offering enhanced robustness against deletion operations. Through extensive experiments on two public datasets, the experimental results demonstrate that our approach has high detection performance and robustness against different adversarial attacks.

Sequence-based deep learning models (e.g., RNNs), can detect malware by analyzing its behavioral sequences. Meanwhile, these models are susceptible to adversarial attacks. Attackers can create adversarial samples that alter the sequence characteristics of behavior sequences to deceive malware classifiers. The existing methods for generating adversarial samples typically involve deleting or replacing crucial behaviors in the original data sequences, or inserting benign behaviors that may violate the behavior constraints. However, these methods that directly manipulate sequences make adversarial samples difficult to implement or apply in practice. In this paper, we propose an adversarial attack approach based on Deep Q-Network and a heuristic backtracking search strategy, which can generate perturbation sequences that satisfy practical conditions for successful attacks. Subsequently, we utilize a novel transformation approach that maps modifications back to the source code, thereby avoiding the need to directly modify the behavior log sequences. We conduct an evaluation of our approach, and the results confirm its effectiveness in generating adversarial samples from real-world malware behavior sequences, which have a high success rate in evading anomaly detection models. Furthermore, our approach is practical and can generate adversarial samples while maintaining the functionality of the modified software.

With the development of Internet of Things (IoT), it is gaining a lot of attention. It is important to secure the embedded systems with low overhead. The Linux Seccomp is widely used by developers to secure the kernels by blocking the access of unused syscalls, which introduces less overhead. However, there are no systematic Seccomp configuration approaches for IoT applications without the help of developers. In addition, the existing Seccomp configuration approaches are coarse-grained, which cannot analyze and limit the syscall arguments. In this article, a novel static dependent syscall analysis approach for embedded applications is proposed, which can obtain all of the possible dependent syscalls and the corresponding arguments of the target applications. So, a fine-grained kernel access limitation can be performed for the IoT applications. To this end, the mappings between dynamic library APIs and syscalls according with their arguments are built, by analyzing the control flow graphs and the data dependency relationships of the dynamic libraries. To the best of our knowledge, this is the first work to generate the fine-grained Seccomp profile for embedded applications.

Sequential deep learning models (e.g., RNN and LSTM) can learn the sequence features of software behaviors, such as API or syscall sequences. However, recent studies have shown that these deep learning-based approaches are vulnerable to adversarial samples. Attackers can use adversarial samples to change the sequential characteristics of behavior sequences and mislead malware classifiers. In this paper, an adversarial robustness anomaly detection method based on the analysis of behavior units is proposed to overcome this problem. We extract related behaviors that usually perform a behavior intention as a behavior unit, which contains the representative semantic information of local behaviors and can be used to improve the robustness of behavior analysis. By learning the overall semantics of each behavior unit and the contextual relationships among behavior units based on a multilevel deep learning model, our approach can mitigate perturbation attacks that target local and large-scale behaviors. In addition, our approach can be applied to both low-level and high-level behavior logs (e.g., API and syscall logs). The experimental results show that our approach outperforms all the compared methods, which indicates that our approach has better performance against obfuscation attacks.

Linux Seccomp is widely used by the program developers and the system maintainers to secure the operating systems, which can block unused syscalls for different applications and containers to shrink the attack surface of the operating systems. However, it is difficult to configure the whitelist of a container or application without the help of program developers. Docker containers block about only 50 syscalls by default, and lots of unblocked useless syscalls introduce a big kernel attack surface. To obtain the dependent syscalls, dynamic tracking is a straight-forward approach but it cannot get the full syscall list. Static analysis can construct an over-approximated syscall list, but the list contains many false positives. In this paper, a systematic dependent syscall analysis approach, sysverify, is proposed by combining static analysis and dynamic verification together to shrink the kernel attack surface. The semantic gap between the binary executables and syscalls is bridged by analyzing the binary and the source code, which builds the mapping between the library APIs and syscalls systematically. To further reduce the attack surface at best effort, we propose a dynamic verification approach to intercept and analyze the security of the invocations of indirect-call-related or rarely invoked syscalls with low overhead.

Error handling is essential for operating systems, thus, there are many bugs in error-handling code, which could result in serious consequences. In this paper, we revisit the problem of error miss-handling bugs and analyze the root cause of the most common ones in the Linux kernel. Based on the analysis, we propose a systematic static taint-analysis-based approach, ErrHunter, to detect multiple kinds of error miss-handling bugs in the Linux kernel. An automated critical variable identification approach is proposed to identify critical variables in the error-handling paths. A static cross-control-flow taint analysis approach is proposed to construct critical-variable control flow graphs (CCFGs), which describe the processing of critical variables in separate control flows. Based on the CCFGs, ErrHunter can target the root cause of the most common error miss-handling bugs and detect the bugs in a systematic way. ErrHunter is designed for kernel bug detection, so it can handle many specific features of the Linux kernel, such as memory management mechanisms, etc.

Monitoring kernel object modification of virtual machine is widely used by virtual-machine-introspection-based security monitors to protect virtual machines in cloud computing, such as monitoring dentry objects to intercept file operations, etc. However, most of the current virtual machine monitors, such as KVM and Xen, only support page-level monitoring, because the Intel EPT technology can only monitor page privilege. If the out-of-virtual-machine security tools want to monitor some kernel objects, they need to intercept the operation of the whole memory page. Since there are some other objects stored in the monitored pages, the modification of them will also trigger the monitor. Therefore, page-level memory monitor usually introduces overhead to related kernel services of the target virtual machine. In this paper, we propose a low-overhead kernel object monitoring approach to reduce the overhead caused by page-level monitor. The core idea is to migrate the target kernel objects to a protected memory area and then to monitor the corresponding new memory pages. Since the new pages only contain the kernel objects to be monitored, other kernel objects will not trigger our monitor. Therefore, our monitor will not introduce runtime overhead to the related kernel service. The experimental results show that our system can monitor target kernel objects effectively only with very low overhead.

Summary In‐VM management is usually needed by cloud service providers for cloud management, which includes monitoring the in‐VM application running state, reconfiguring VM system settings, etc. In‐VM management is also very useful in green cloud computing, because it provides the abilities of in‐VM monitoring, VM reconfiguration, performance measurement, etc. Leveraging a shell or an in‐VM agent to manage VMs is faced with generality and security challenges. In this paper, we propose a secure automated in‐VM management approach, ie, SAVM, which likes a hypervisor‐based shell managing the VMs in an out‐of‐box way. To bridge the semantic gap, we reuse the target VM's system calls to process the semantic information automatically. More importantly, we introduce a secure instruction fetch approach to enhance the system security. As a result, SAVM does not rely on the target VM's kernel integrity. In addition, we also present a dummy process selection and a system call injection method to further enhance the system security and transparency. After the implementation, we evaluate the prototype. The experimental results show that SAVM can achieve most of the in‐VM management operations. Furthermore, SAVM can work correctly under the target VM attacked by several popular rootkits.

Protecting critical files in file systems is very important to computer systems. To protect critical files, the VMI-based Real-time File-system Monitor tools are promising options. However, these tools are always operation-based and introduce high overhead. The operation-based approaches intercept some kind of file operation to monitor critical files. The selected file operation is intercepted by the monitor whenever it is being executed. As file operation are high-frequency, the operation-based methods always result in the high performance degradation. In this paper, we present a VMI-based low overhead real-time critical file monitor method, CFWatcher, to meet the performance requirements of real-time monitor tools. CFWatcher is a target-based monitor tool which means it only intercepts the file operations accessing the user-defined critical files, and then obtains enough information to check the rules. The overhead of CFWatcher is related to the frequency of the target being accessed. Besides monitoring critical files, CFWatcher can take actions to prevent the illegal access if there is any rule violation. We implemented the prototype of CFWatcher and then evaluated the performance. Experimental results show that the overhead of our approach is low.