arXiv subjects

Zhili Zhou

Publications and source records attributed to Zhili Zhou.

17 recordsLinked to original sources

arXiv2025

Beyond Statistical Estimation: Differentially Private Individual Computation via Shuffling

In data-driven applications, preserving user privacy while enabling valuable computations remains a critical challenge. Technologies like differential privacy have been pivotal in addressing these concerns. The shuffle model of DP requires no trusted curators and can achieve high utility by leveraging the privacy amplification effect yielded from shuffling. These benefits have led to significant interest in the shuffle model. However, the computation tasks in the shuffle model are limited to statistical estimation, making it inapplicable to real-world scenarios in which each user requires a personalized output. This paper introduces a novel paradigm termed Private Individual Computation (PIC), expanding the shuffle model to support a broader range of permutation-equivariant computations. PIC enables personalized outputs while preserving privacy, and enjoys privacy amplification through shuffling. We propose a concrete protocol that realizes PIC. By using one-time public keys, our protocol enables users to receive their outputs without compromising anonymity, which is essential for privacy amplification. Additionally, we present an optimal randomizer, the Minkowski Response, designed for the PIC model to enhance utility. We formally prove the security and privacy properties of the PIC protocol. Theoretical analysis and empirical evaluations demonstrate PIC's capability in handling non-statistical computation tasks, and the efficacy of PIC and the Minkowski randomizer in achieving superior utility compared to existing solutions.

cs.CR↗

arXiv2025

SGC-Net: Stratified Granular Comparison Network for Open-Vocabulary HOI Detection

Recent open-vocabulary human-object interaction (OV-HOI) detection methods primarily rely on large language model (LLM) for generating auxiliary descriptions and leverage knowledge distilled from CLIP to detect unseen interaction categories. Despite their effectiveness, these methods face two challenges: (1) feature granularity deficiency, due to reliance on last layer visual features for text alignment, leading to the neglect of crucial object-level details from intermediate layers; (2) semantic similarity confusion, resulting from CLIP's inherent biases toward certain classes, while LLM-generated descriptions based solely on labels fail to adequately capture inter-class similarities. To address these challenges, we propose a stratified granular comparison network. First, we introduce a granularity sensing alignment module that aggregates global semantic features with local details, refining interaction representations and ensuring robust alignment between intermediate visual features and text embeddings. Second, we develop a hierarchical group comparison module that recursively compares and groups classes using LLMs, generating fine-grained and discriminative descriptions for each interaction category. Experimental results on two widely-used benchmark datasets, SWIG-HOI and HICO-DET, demonstrate that our method achieves state-of-the-art results in OV-HOI detection. Codes will be released on https://github.com/Phil0212/SGC-Net.

cs.CV↗

arXiv2025

Refining Positive and Toxic Samples for Dual Safety Self-Alignment of LLMs with Minimal Human Interventions

Recent AI agents, such as ChatGPT and LLaMA, primarily rely on instruction tuning and reinforcement learning to calibrate the output of large language models (LLMs) with human intentions, ensuring the outputs are harmless and helpful. Existing methods heavily depend on the manual annotation of high-quality positive samples, while contending with issues such as noisy labels and minimal distinctions between preferred and dispreferred response data. However, readily available toxic samples with clear safety distinctions are often filtered out, removing valuable negative references that could aid LLMs in safety alignment. In response, we propose PT-ALIGN, a novel safety self-alignment approach that minimizes human supervision by automatically refining positive and toxic samples and performing fine-grained dual instruction tuning. Positive samples are harmless responses, while toxic samples deliberately contain extremely harmful content, serving as a new supervisory signals. Specifically, we utilize LLM itself to iteratively generate and refine training instances by only exploring fewer than 50 human annotations. We then employ two losses, i.e., maximum likelihood estimation (MLE) and fine-grained unlikelihood training (UT), to jointly learn to enhance the LLM's safety. The MLE loss encourages an LLM to maximize the generation of harmless content based on positive samples. Conversely, the fine-grained UT loss guides the LLM to minimize the output of harmful words based on negative samples at the token-level, thereby guiding the model to decouple safety from effectiveness, directing it toward safer fine-tuning objectives, and increasing the likelihood of generating helpful and reliable content. Experiments on 9 popular open-source LLMs demonstrate the effectiveness of our PT-ALIGN for safety alignment, while maintaining comparable levels of helpfulness and usefulness.

cs.CL↗

arXiv2024

Efficient Streaming Voice Steganalysis in Challenging Detection Scenarios

In recent years, there has been an increasing number of information hiding techniques based on network streaming media, focusing on how to covertly and efficiently embed secret information into real-time transmitted network media signals to achieve concealed communication. The misuse of these techniques can lead to significant security risks, such as the spread of malicious code, commands, and viruses. Current steganalysis methods for network voice streams face two major challenges: efficient detection under low embedding rates and short duration conditions. These challenges arise because, with low embedding rates (e.g., as low as 10%) and short transmission durations (e.g., only 0.1 second), detection models struggle to acquire sufficiently rich sample features, making effective steganalysis difficult. To address these challenges, this paper introduces a Dual-View VoIP Steganalysis Framework (DVSF). The framework first randomly obfuscates parts of the native steganographic descriptors in VoIP stream segments, making the steganographic features of hard-to-detect samples more pronounced and easier to learn. It then captures fine-grained local features related to steganography, building on the global features of VoIP. Specially constructed VoIP segment triplets further adjust the feature distances within the model. Ultimately, this method effectively address the detection difficulty in VoIP. Extensive experiments demonstrate that our method significantly improves the accuracy of streaming voice steganalysis in these challenging detection scenarios, surpassing existing state-of-the-art methods and offering superior near-real-time performance.

cs.CR↗

arXiv2024

ContextBLIP: Doubly Contextual Alignment for Contrastive Image Retrieval from Linguistically Complex Descriptions

Image retrieval from contextual descriptions (IRCD) aims to identify an image within a set of minimally contrastive candidates based on linguistically complex text. Despite the success of VLMs, they still significantly lag behind human performance in IRCD. The main challenges lie in aligning key contextual cues in two modalities, where these subtle cues are concealed in tiny areas of multiple contrastive images and within the complex linguistics of textual descriptions. This motivates us to propose ContextBLIP, a simple yet effective method that relies on a doubly contextual alignment scheme for challenging IRCD. Specifically, 1) our model comprises a multi-scale adapter, a matching loss, and a text-guided masking loss. The adapter learns to capture fine-grained visual cues. The two losses enable iterative supervision for the adapter, gradually highlighting the focal patches of a single image to the key textual cues. We term such a way as intra-contextual alignment. 2) Then, ContextBLIP further employs an inter-context encoder to learn dependencies among candidates, facilitating alignment between the text to multiple images. We term this step as inter-contextual alignment. Consequently, the nuanced cues concealed in each modality can be effectively aligned. Experiments on two benchmarks show the superiority of our method. We observe that ContextBLIP can yield comparable results with GPT-4V, despite involving about 7,500 times fewer parameters.

cs.CV↗

arXiv2024

A Dynamic YOLO-Based Sequence-Matching Model for Efficient Coverless Image Steganography

Many existing coverless steganography methods establish a mapping relationship between cover images and hidden data. There exists an issue that the number of images stored in the database grows exponentially as the steganographic capacity rises. The need for a high steganographic capacity makes it challenging to build an image database. To improve the image library utilization and anti-attack capability of the steganography system, we present an efficient coverless scheme based on dynamically matched substrings. YOLO is employed for selecting optimal objects, and a mapping dictionary is established between these objects and scrambling factors. With the aid of this dictionary, each image is effectively assigned to a specific scrambling factor, which is used to scramble the receiver's sequence key. To achieve sufficient steganography capability based on a limited image library, all substrings of the scrambled sequences hold the potential to hide data. After completing the secret information matching, the ideal number of stego images will be obtained from the database. According to experimental results, this technology outperforms most previous works on data load, transmission security, and hiding capacity. Under typical geometric attacks, it can recover 79.85\% of secret information on average. Furthermore, only approximately 200 random images are needed to meet a capacity of 19 bits per image.

cs.CR↗

arXiv2023

GanFinger: GAN-Based Fingerprint Generation for Deep Neural Network Ownership Verification

Deep neural networks (DNNs) are extensively employed in a wide range of application scenarios. Generally, training a commercially viable neural network requires significant amounts of data and computing resources, and it is easy for unauthorized users to use the networks illegally. Therefore, network ownership verification has become one of the most crucial steps in safeguarding digital assets. To verify the ownership of networks, the existing network fingerprinting approaches perform poorly in the aspects of efficiency, stealthiness, and discriminability. To address these issues, we propose a network fingerprinting approach, named as GanFinger, to construct the network fingerprints based on the network behavior, which is characterized by network outputs of pairs of original examples and conferrable adversarial examples. Specifically, GanFinger leverages Generative Adversarial Networks (GANs) to effectively generate conferrable adversarial examples with imperceptible perturbations. These examples can exhibit identical outputs on copyrighted and pirated networks while producing different results on irrelevant networks. Moreover, to enhance the accuracy of fingerprint ownership verification, the network similarity is computed based on the accuracy-robustness distance of fingerprint examples'outputs. To evaluate the performance of GanFinger, we construct a comprehensive benchmark consisting of 186 networks with five network structures and four popular network post-processing techniques. The benchmark experiments demonstrate that GanFinger significantly outperforms the state-of-the-arts in efficiency, stealthiness, and discriminability. It achieves a remarkable 6.57 times faster in fingerprint generation and boosts the ARUC value by 0.175, resulting in a relative improvement of about 26%.

cs.CR↗

arXiv2023

Real-World Image Super Resolution via Unsupervised Bi-directional Cycle Domain Transfer Learning based Generative Adversarial Network

Deep Convolutional Neural Networks (DCNNs) have exhibited impressive performance on image super-resolution tasks. However, these deep learning-based super-resolution methods perform poorly in real-world super-resolution tasks, where the paired high-resolution and low-resolution images are unavailable and the low-resolution images are degraded by complicated and unknown kernels. To break these limitations, we propose the Unsupervised Bi-directional Cycle Domain Transfer Learning-based Generative Adversarial Network (UBCDTL-GAN), which consists of an Unsupervised Bi-directional Cycle Domain Transfer Network (UBCDTN) and the Semantic Encoder guided Super Resolution Network (SESRN). First, the UBCDTN is able to produce an approximated real-like LR image through transferring the LR image from an artificially degraded domain to the real-world LR image domain. Second, the SESRN has the ability to super-resolve the approximated real-like LR image to a photo-realistic HR image. Extensive experiments on unpaired real-world image benchmark datasets demonstrate that the proposed method achieves superior performance compared to state-of-the-art methods.

cs.CV↗

arXiv2022

Auto-Focus Contrastive Learning for Image Manipulation Detection

Generally, current image manipulation detection models are simply built on manipulation traces. However, we argue that those models achieve sub-optimal detection performance as it tends to: 1) distinguish the manipulation traces from a lot of noisy information within the entire image, and 2) ignore the trace relations among the pixels of each manipulated region and its surroundings. To overcome these limitations, we propose an Auto-Focus Contrastive Learning (AF-CL) network for image manipulation detection. It contains two main ideas, i.e., multi-scale view generation (MSVG) and trace relation modeling (TRM). Specifically, MSVG aims to generate a pair of views, each of which contains the manipulated region and its surroundings at a different scale, while TRM plays a role in modeling the trace relations among the pixels of each manipulated region and its surroundings for learning the discriminative representation. After learning the AF-CL network by minimizing the distance between the representations of corresponding views, the learned network is able to automatically focus on the manipulated region and its surroundings and sufficiently explore their trace relations for accurate manipulation detection. Extensive experiments demonstrate that, compared to the state-of-the-arts, AF-CL provides significant performance improvements, i.e., up to 2.5%, 7.5%, and 0.8% F1 score, on CAISA, NIST, and Coverage datasets, respectively.

cs.CV↗

arXiv2022

Traceable and Authenticable Image Tagging for Fake News Detection

To prevent fake news images from misleading the public, it is desirable not only to verify the authenticity of news images but also to trace the source of fake news, so as to provide a complete forensic chain for reliable fake news detection. To simultaneously achieve the goals of authenticity verification and source tracing, we propose a traceable and authenticable image tagging approach that is based on a design of Decoupled Invertible Neural Network (DINN). The designed DINN can simultaneously embed the dual-tags, \textit{i.e.}, authenticable tag and traceable tag, into each news image before publishing, and then separately extract them for authenticity verification and source tracing. Moreover, to improve the accuracy of dual-tags extraction, we design a parallel Feature Aware Projection Model (FAPM) to help the DINN preserve essential tag information. In addition, we define a Distance Metric-Guided Module (DMGM) that learns asymmetric one-class representations to enable the dual-tags to achieve different robustness performances under malicious manipulations. Extensive experiments, on diverse datasets and unseen manipulations, demonstrate that the proposed tagging approach achieves excellent performance in the aspects of both authenticity verification and source tracing for reliable fake news detection and outperforms the prior works.

cs.CV↗

arXiv2022

Continual Learning for Steganalysis

To detect the existing steganographic algorithms, recent steganalysis methods usually train a Convolutional Neural Network (CNN) model on the dataset consisting of corresponding paired cover/stego-images. However, it is inefficient and impractical for those steganalysis tools to completely retrain the CNN model to make it effective against both the existing steganographic algorithms and a new emerging steganographic algorithm. Thus, existing steganalysis models usually lack dynamic extensibility for new steganographic algorithms, which limits their application in real-world scenarios. To address this issue, we propose an accurate parameter importance estimation (APIE) based-continual learning scheme for steganalysis. In this scheme, when a steganalysis model is trained on the new image dataset generated by the new steganographic algorithm, its network parameters are effectively and efficiently updated with sufficient consideration of their importance evaluated in the previous training process. This approach can guide the steganalysis model to learn the patterns of the new steganographic algorithm without significantly degrading the detectability against the previous steganographic algorithms. Experimental results demonstrate the proposed scheme has promising extensibility for new emerging steganographic algorithms.

cs.CV↗

arXiv2022

Secret-to-Image Reversible Transformation for Generative Steganography

Recently, generative steganography that transforms secret information to a generated image has been a promising technique to resist steganalysis detection. However, due to the inefficiency and irreversibility of the secret-to-image transformation, it is hard to find a good trade-off between the information hiding capacity and extraction accuracy. To address this issue, we propose a secret-to-image reversible transformation (S2IRT) scheme for generative steganography. The proposed S2IRT scheme is based on a generative model, i.e., Glow model, which enables a bijective-mapping between latent space with multivariate Gaussian distribution and image space with a complex distribution. In the process of S2I transformation, guided by a given secret message, we construct a latent vector and then map it to a generated image by the Glow model, so that the secret message is finally transformed to the generated image. Owing to good efficiency and reversibility of S2IRT scheme, the proposed steganographic approach achieves both high hiding capacity and accurate extraction of secret message from generated image. Furthermore, a separate encoding-based S2IRT (SE-S2IRT) scheme is also proposed to improve the robustness to common image attacks. The experiments demonstrate the proposed steganographic approaches can achieve high hiding capacity (up to 4 bpp) and accurate information extraction (almost 100% accuracy rate) simultaneously, while maintaining desirable anti-detectability and imperceptibility.

cs.CR↗

arXiv2022

Learning Hierarchical Graph Representation for Image Manipulation Detection

The objective of image manipulation detection is to identify and locate the manipulated regions in the images. Recent approaches mostly adopt the sophisticated Convolutional Neural Networks (CNNs) to capture the tampering artifacts left in the images to locate the manipulated regions. However, these approaches ignore the feature correlations, i.e., feature inconsistencies, between manipulated regions and non-manipulated regions, leading to inferior detection performance. To address this issue, we propose a hierarchical Graph Convolutional Network (HGCN-Net), which consists of two parallel branches: the backbone network branch and the hierarchical graph representation learning (HGRL) branch for image manipulation detection. Specifically, the feature maps of a given image are extracted by the backbone network branch, and then the feature correlations within the feature maps are modeled as a set of fully-connected graphs for learning the hierarchical graph representation by the HGRL branch. The learned hierarchical graph representation can sufficiently capture the feature correlations across different scales, and thus it provides high discriminability for distinguishing manipulated and non-manipulated regions. Extensive experiments on four public datasets demonstrate that the proposed HGCN-Net not only provides promising detection accuracy, but also achieves strong robustness under a variety of common image attacks in the task of image manipulation detection, compared to the state-of-the-arts.

cs.CV↗

arXiv2019

Robust Network Function Virtualization

Network function virtualization (NFV) enables on-demand network function (NF) deployment providing agile and dynamic network services. Through an evaluation metric that quantifies the minimal reliability among all NFs for all demands, service providers and operators may better facilitate flexible NF service recovery and migration, thus offer higher service reliability. In this paper, we present evaluation metrics on NFV reliability and solution approaches to solve robust NFV under random NF-enabled node failure(s). We demonstrate how to construct an auxiliary NF-enabled network and its mapping onto the physical substrate network. With constructed NF-enabled network, we develop pseudo-polynomial algorithms to solve the robust NF and SFC $s-t$ path problems -- subproblems of robust NFV. We also present approximation algorithms for robust NFV with the SFC-Fork as the NF forwarding graph. Furthermore, we propose exact solution approaches via mixed-integer linear programming (MILP) under the general setting. Computational results show that our proposed solution approaches are capable of managing robust NFV in a large-size network.

cs.NI↗

arXiv2017

Survivable Probability of Network Slicing with Random Physical Link Failure

The fifth generation of communication technology (5G) revolutionizes mobile networks and the associated ecosystems through the integration of cross-domain networks. Network slicing is an enabling technology for 5G as it provides dynamic, on-demand, and reliable logical network slices (i.e., network services) over a common physical network/infrastructure. Since a network slice is subject to failures originated from disruptions, namely node or link failures, in the physical infrastructure, our utmost interest is to evaluate the reliability of a network slice before assigning it to customers. In this paper, we propose an evaluation metric, \textit{survivable probability}, to quantify the reliability of a network slice under random physical link failure(s). We prove the existence of a \textit{base protecting spanning tree set} which has the same survivable probability as that of a network slice. We propose the necessary and sufficient conditions to identify a base protecting spanning tree set and develop corresponding mathematical formulations, which can be used to generate reliable network slices in the 5G environment. In addition to proving the viability of our approaches with simulation results, we also discuss how our problems and approaches are related to the Steiner tree problems and present their computational complexity and approximability.

cs.NI↗

arXiv2017

Survivable Probability of SDN-enabled Cloud Networking with Random Physical Link Failure

Software-driven cloud networking is a new paradigm in orchestrating physical resources (CPU, network bandwidth, energy, storage) allocated to network functions, services, and applications, which is commonly modeled as a cross-layer network. This model carries a physical network representing the physical infrastructure, a logical network showing demands, and logical-to-physical node/link mappings. In such networks, a single failure in the physical network may trigger cascading failures in the logical network and disable network services and connectivity. In this paper, we propose an evaluation metric, survivable probability, to evaluate the reliability of such networks under random physical link failure(s). We propose the concept of base protecting spanning tree and prove the necessary and sufficient conditions for its existence and relation to survivability. We then develop mathematical programming formulations for reliable cross-layer network routing design with the maximal reliable probability. Computation results demonstrate the viability of our approach.

cs.NI↗

arXiv2016

Survivable Cloud Network Design Against Multiple Failures Through Protecting Spanning Trees

Survivable design of cross-layer networks, such as the cloud computing infrastructure, lies in its resource deployment and allocation and mapping of the logical (virtual datacenter/IP) network into the physical infrastructure (cloud backbone/WDM) such that link or node failure(s) in the physical infrastructure would not result in cascading failures in the logical network. Most of the prior approaches for survivable cross-layer network design aim at single-link failure scenario, which are not applicable to the more challenging multi-failure scenarios. Also, as many of these approaches use the cross-layer cut concept, enumeration of all cuts in the network is required and thus introducing exponential number of constraints. To overcome these difficulties, we investigate in this paper survivable mapping approaches against multiple physical link failures and its special case, Shared Risk Link Group (SRLG) failure. We present the necessary and sufficient conditions based on both cross-layer spanning trees and cutsets to guarantee a survivable mapping when multiple physical link failures occur. Based on the necessary and sufficient conditions, we propose to solve the problem through (1) mixed-integer linear programs which avoid enumerating all combinations of link failures, and (2) an algorithm which generates/adds logical spanning trees sequentially. Our simulation results show that the proposed approaches can produce survivable mappings effectively against both $k$- and SRLG-failures.

cs.NI↗