Author name cluster

Hui Wei

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

9 papers

2 author rows

AAAI Conference 2025 Conference Paper

Balancing Privacy and Performance: A Many-in-One Approach for Image Anonymization

Xuemei Jia
Jiawei Du
Hui Wei
Ruinian Xue
Zheng Wang
Hongyuan Zhu
Jun Chen

The effective utilization of data through Deep Neural Networks (DNNs) has profoundly influenced various aspects of society. The growing demand for high-quality, particularly personalized, data has spurred research efforts to prevent data leakage and protect privacy in recent years. Early privacy-preserving methods primarily relied on instance-wise modifications, such as erasing or obfuscating essential features for de-identification. However, this approach highlights an inherent trade-off: minimal modification offers insufficient privacy protection, while excessive modification significantly degrades task performance. In this paper, we propose a novel Recombining for Obfuscation (FRO) approach to address this trade-off. Unlike existing methods that generate one anonymized instance by perturbing the original data on a one-to-one basis, our FRO approach generates an anonymized instance by reassembling mixed ID-related features from multiple original data sources on a many-in-one basis. Instead of introducing additional noise for de-identification, our approach leverages the existing non-polluted features from other instances to anonymize data. Extensive experiments on identity identification tasks demonstrate that FRO outperforms previous state-of-the-art methods, not only in utility performance but also in visual anonymization.

PDF Details DOI

AAAI Conference 2025 Conference Paper

MI-CAPTCHA: Enhance the Security of CAPTCHA Using Mooney Images

Jingmeng Li
Lukang Fu
Surun Yang
Hui Wei

Completely automated public Turing test to tell humans apart (CAPTCHA) is an effective mechanism to protect websites and online applications from malicious bots programs. Image-based CAPTCHA is one of the most widely used schemes. However, deep learning techniques have significantly weakened the security of some image-based CAPTCHA schemes. Mooney images (MIs) are important research materials in the field of cognitive science. Compared to natural images, MI exhibits fewer visual cues, fragmented content, and greater ambiguity, leading to the perception of MI relying more on the iterative process between feedforward and feedback mechanisms. In this paper, we raise an intriguing question: can MIs be used to enhance the security of CAPTCHA? Before this study, we first propose a novel framework HiMI that generates the high-quality MIs from natural images and also allows flexible adjustment of the perceived difficulty. Based on MI, we design two MI-CAPTCHA schemes related to object detection and instance segmentation tasks, respectively. We experimentally demonstrate that HiMI performs better than other baseline methods in terms of both image quality and application potential in two MI-CAPTCHA schemes. Additionally, we conduct experiments to explore the solving performance of humans and CAPTCHA solvers under different parameter settings of schemes, providing valuable reference for the practical application.

PDF Details DOI

ECAI Conference 2025 Conference Paper

SpectraShift3D: Black-Box Physical Attacks on Monocular Depth Estimation via Robust 3D-Aware Textures

Changda Shi
Lili Zhou
Hui Wei

Monocular Depth Estimation (MDE) is a core component in safety-critical systems such as autonomous driving. However, its vulnerability to physical adversarial attacks under practical black-box settings—where attackers cannot access model parameters—has not been thoroughly investigated. This paper proposes SpectraShift3D, a novel black-box physical attack framework that generates 3D-perceptual adversarial textures without requiring any model access. The framework integrates three key innovations: (1) a Feature Consistency Loss to enhance transferability across diverse MDE architectures; (2) an Adaptive Dynamic Depth Error Loss that selectively maximizes depth distortion in safety-critical regions (e. g. , road obstacles); and (3) a Dynamic Occlusion Shape module to generate physically plausible occlusions for real-world robustness. Extensive experiments on six state-of-the-art MDE models (including MonoDepth2, MiDaS, and DepthAnything) show that SpectraShift3D misleads depth estimation by 13. 31 meters on average (surpassing safe driving thresholds) with a 53. 6% attack success rate, significantly outperforming existing methods. The attack remains effective under varying viewpoints, lighting, and distances. Our work highlights critical security risks in real-world MDE systems and establishes a standardized benchmark for black-box physical attacks, urging the community to reconsider the robustness of vision-based autonomous systems.

Details

IJCAI Conference 2024 Conference Paper

A New Guaranteed Outlier Removal Method Based on Plane Constraints for Large-Scale LiDAR Point Cloud Registration

Gang Ma
Hui Wei
Runfeng Lin
Jialiang Wu

In this paper, we present a novel registration method based on plane constraints for large-scale LiDAR point clouds, effectively decoupling rotation estimation and translation estimation. For rotation estimation, we propose an outlier removal method that combines coarse filtering with rotation-invariant constraints and refined filtering based on computational geometric consistency checks, effectively pruning outliers and robustly estimating accurate relative rotations from plane normals. In translation estimation, we propose a component-wise method based on plane translation constraints to efficiently estimate relative translations. The robustness and effectiveness of our proposed method are empirically validated on three popular LiDAR point cloud datasets. The experimental results convincingly demonstrate that our approach achieves state-of-the-art performance.

PDF Details DOI

ICLR Conference 2024 Conference Paper

REBAR: Retrieval-Based Reconstruction for Time-series Contrastive Learning

Maxwell A. Xu
Alexander Moreno
Hui Wei
Benjamin M. Marlin
James M. Rehg

The success of self-supervised contrastive learning hinges on identifying positive data pairs, such that when they are pushed together in embedding space, the space encodes useful information for subsequent downstream tasks. Constructing positive pairs is non-trivial as the pairing must be similar enough to reflect a shared semantic meaning, but different enough to capture within-class variation. Classical approaches in vision use augmentations to exploit well-established invariances to construct positive pairs, but invariances in the time-series domain are much less obvious. In our work, we propose a novel method of using a learned measure for identifying positive pairs. Our Retrieval-Based Reconstruction (REBAR) measure measures the similarity between two sequences as the reconstruction error that results from reconstructing one sequence with retrieved information from the other. Then, if the two sequences have high REBAR similarity, we label them as a positive pair. Through validation experiments, we show that the REBAR error is a predictor of mutual class membership. Once integrated into a contrastive learning framework, our REBAR method learns an embedding that achieves state-of-the-art performance on downstream tasks across various modalities.

Details

NeurIPS Conference 2024 Conference Paper

Revisiting Adversarial Patches for Designing Camera-Agnostic Attacks against Person Detection

Hui Wei
Zhixiang Wang
Kewei Zhang
Jiaqi Hou
Yuanwei Liu
Hao Tang
Zheng Wang

Physical adversarial attacks can deceive deep neural networks (DNNs), leading to erroneous predictions in real-world scenarios. To uncover potential security risks, attacking the safety-critical task of person detection has garnered significant attention. However, we observe that existing attack methods overlook the pivotal role of the camera, involving capturing real-world scenes and converting them into digital images, in the physical adversarial attack workflow. This oversight leads to instability and challenges in reproducing these attacks. In this work, we revisit patch-based attacks against person detectors and introduce a camera-agnostic physical adversarial attack to mitigate this limitation. Specifically, we construct a differentiable camera Image Signal Processing (ISP) proxy network to compensate for the physical-to-digital transition gap. Furthermore, the camera ISP proxy network serves as a defense module, forming an adversarial optimization framework with the attack module. The attack module optimizes adversarial patches to maximize effectiveness, while the defense module optimizes the conditional parameters of the camera ISP proxy network to minimize attack effectiveness. These modules engage in an adversarial game, enhancing cross-camera stability. Experimental results demonstrate that our proposed Camera-Agnostic Patch (CAP) attack effectively conceals persons from detectors across various imaging hardware, including two distinct cameras and four smartphones.

PDF Details DOI

AAAI Conference 2023 Conference Paper

HOTCOLD Block: Fooling Thermal Infrared Detectors with a Novel Wearable Design

Hui Wei
Zhixiang Wang
Xuemei Jia
Yinqiang Zheng
Hao Tang
Shin'ichi Satoh
Zheng Wang

Adversarial attacks on thermal infrared imaging expose the risk of related applications. Estimating the security of these systems is essential for safely deploying them in the real world. In many cases, realizing the attacks in the physical space requires elaborate special perturbations. These solutions are often impractical and attention-grabbing. To address the need for a physically practical and stealthy adversarial attack, we introduce HotCold Block, a novel physical attack for infrared detectors that hide persons utilizing the wearable Warming Paste and Cooling Paste. By attaching these readily available temperature-controlled materials to the body, HotCold Block evades human eyes efficiently. Moreover, unlike existing methods that build adversarial patches with complex texture and structure features, HotCold Block utilizes an SSP-oriented adversarial optimization algorithm that enables attacks with pure color blocks and explores the influence of size, shape, and position on attack performance. Extensive experimental results in both digital and physical environments demonstrate the performance of our proposed HotCold Block. Code is available: https://github.com/weihui1308/HOTCOLDBlock.

PDF Details DOI

IS Journal 2022 Journal Article

Recognizing Slanted Deck Scenes by Non-Manhattan Spatial Right Angle Projection

Luping Wang
Hui Wei

Recognizing slanted deck scenes is crucial to security monitoring for protecting ships and making them behave smartly resilient. However, there are multitude of diverse structures that are designed as slanted planes due to rough maritime environments. Traditional methods for scene understanding from 3D point clouds or RGB-D data are energy-consuming and memory intensive, which makes those models less reliable in a resource-constrained system of limited compute, memory, and energy resources on ships. In this study, we present an approach to understanding deck scenes, including slanted structures, using a low-cost monocular camera without prior training. New clusters of slanted angle projections are extracted. The vanishing points of slanted non-Manhattan angle projections are estimated. These slanted planes can be reshaped by compositions of non-Manhattan angle projections. Combined with Manhattan planes, a deck scene can be approximated by Manhattan and non-Manhattan planes. Unlike deep learning-based algorithms, this approach requires no prior training or knowledge of the camera’s internal parameters. Experimental results demonstrated that the method can successfully elucidate diverse elements, including slanted structures, meeting safety monitoring requirements using a resource-constrained monocular camera in a deck environment.

Details DOI

EAAI Journal 2020 Journal Article

Understanding of wheelchair ramp scenes for disabled people with visual impairments

Luping Wang
Hui Wei

Helping disabled people with visual impairments understand wheelchair ramp scenes has considerable value in computer vision. However, due to the diversity of wheelchair ramp scenes, understanding them remains a big challenge. Wheelchair ramp planes can be considered as a composition of rectangles that do not satisfy manhattan assumption. These non-manhattan rectangles are projected into two dimensional projections, shaping into special geometric configurations, which may enable us to estimate their original orientation and position in 3D scenes. In this paper, we presented a method for disabled people with visual impairments to understand wheelchair ramp scenes from a single image without any prior training. Firstly angle projections can be assigned to different clusters. Secondly ramp vanishing points (RVPs) can be estimated. Then it is possible to determine ramp planes consisting of angle projections that belong to the estimated RVPs. Finally, the algorithm can understand wheelchair ramp scenes including not only manhattan structures but also ramp planes belonging to non-manhattan structures. The proposed approach requires no prior training or any knowledge of the camera’s internal parameters. Besides, it is robust to the errors in calibration and image noise. We compared the estimated wheelchair ramp scene layout against the ground truth, measuring the percentage of pixels that were incorrectly classified. The experimental results showed that the method can understand wheelchair ramp scenes including not only manhattan structures but also non-manhattan structures of ramp planes, making it practical and efficient for disabled people with visual impairments.

Details DOI