Author name cluster

Feng Lin

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.

14 papers

2 author rows

AAAI Conference 2026 Conference Paper

DySy-Det: A Synergistic Framework with Dynamic Reconstruction-Path Consistency for AI-Generated Image Detection

Fanli Jin
Feng Lin
Gaojian Wang
Tong Wu
Zhisheng Yan

Advanced image generative models have led to concerns about malicious use, underscoring the necessity for generalizable detection methods. However, existing approaches tend to overfit to domain-specific forgery patterns, while overlooking complementary cues from different domains. Therefore, we introduce DySy-Det (Dynamic Synergy Detector), a novel framework that mines collaborative and robust forgery artifacts from multiple evidence domains. First, DySy-Det fine-tunes a CLIP vision transformer to extract high-level semantics for identifying conceptual inconsistencies, while generating attention maps that pinpoint key discriminative regions. Then, this semantic guidance, in the form of a mask, directs a targeted reconstruction process. By focusing on these salient areas, our approach effectively extracts localized reconstruction errors, thereby filtering out irrelevant background noise. Furthermore, inspired by the intrinsic generative mechanics of diffusion models, we introduce the concept of Reconstruction-Path Consistency (RPC), which quantifies the temporal stability of the denoising trajectory to expose dynamic generative artifacts. We capture this by computing noise alignment scores across multiple timesteps and encode them via a lightweight network. Extensive evaluations on GenImage and UniversalFakeDetect benchmarks demonstrate that DySy-Det outperforms the state-of-the-art detector by 6.14% and 1.57% in mean accuracy, respectively.

PDF Details DOI

AAAI Conference 2026 Conference Paper

Eguard: Defending LLM Embeddings Against Inversion Attacks via Text Mutual Information Optimization

Tiantian Liu
Hongwei Yao
Feng Lin
Tong Wu
Zhan Qin
Kui Ren

While text embeddings enable efficient semantic processing in LLMs, they remain vulnerable to inversion attacks that reconstruct sensitive original text. However, current defense methods typically treat text embeddings from the feature level independently, ignoring the exploitation of the mutual relation among the embedding construction pipeline. To address this limitation, we propose Eguard, a framework that effectively disrupts chains of relationships between the original semantic space and defended functional space. Our improvements manifest at two levels, i.e., the global-level and local-level mutual information. At the global level, we propose to minimize the statistical dependency between protected embeddings and their original inputs, effectively decoupling sensitive content from the semantic space accessible to adversaries. At the local level, we apply keyword-antonym contrastive learning to enforce semantic discriminability within the space of downstream utility. This synergy of global privacy control and local semantic alignment allows Eguard to achieve a superior privacy-utility trade-off than traditional defenses. Our approach significantly reduces privacy risks, protecting over 95 percent of tokens from inversion while maintaining high performance across downstream tasks consistent with original embeddings.

PDF Details DOI

AAAI Conference 2024 Conference Paper

Exposing the Deception: Uncovering More Forgery Clues for Deepfake Detection

Zhongjie Ba
Qingyu Liu
Zhenguang Liu
Shuang Wu
Feng Lin
Li Lu
Kui Ren

Deepfake technology has given rise to a spectrum of novel and compelling applications. Unfortunately, the widespread proliferation of high-fidelity fake videos has led to pervasive confusion and deception, shattering our faith that seeing is believing. One aspect that has been overlooked so far is that current deepfake detection approaches may easily fall into the trap of overfitting, focusing only on forgery clues within one or a few local regions. Moreover, existing works heavily rely on neural networks to extract forgery features, lacking theoretical constraints guaranteeing that sufficient forgery clues are extracted and superfluous features are eliminated. These deficiencies culminate in unsatisfactory accuracy and limited generalizability in real-life scenarios. In this paper, we try to tackle these challenges through three designs: (1) We present a novel framework to capture broader forgery clues by extracting multiple non-overlapping local representations and fusing them into a global semantic-rich feature. (2) Based on the information bottleneck theory, we derive Local Information Loss to guarantee the orthogonality of local representations while preserving comprehensive task-relevant information. (3) Further, to fuse the local representations and remove task-irrelevant information, we arrive at a Global Information Loss through the theoretical analysis of mutual information. Empirically, our method achieves state-of-the-art performance on five benchmark datasets. Our code is available at https://github.com/QingyuLiu/Exposing-the-Deception, hoping to inspire researchers.

PDF Details DOI

YNIMG Journal 2024 Journal Article

Identifying individual's distractor suppression using functional connectivity between anatomical large-scale brain regions

Lei Zhuo
Zhenlan Jin
Ke Xie
Simeng Li
Feng Lin
Junjun Zhang
Ling Li

Distractor suppression (DS) is crucial in goal-oriented behaviors, referring to the ability to suppress irrelevant information. Current evidence points to the prefrontal cortex as an origin region of DS, while subcortical, occipital, and temporal regions are also implicated. The present study aimed to examine the contribution of communications between these brain regions to visual DS. To do it, we recruited two independent cohorts of participants for the study. One cohort participated in a visual search experiment where a salient distractor triggering distractor suppression to measure their DS and the other cohort filled out a Cognitive Failure Questionnaire to assess distractibility in daily life. Both cohorts collected resting-state functional magnetic resonance imaging (rs-fMRI) data to investigate function connectivity (FC) underlying DS. First, we generated predictive models of the DS measured in visual search task using resting-state functional connectivity between large anatomical regions. It turned out that the models could successfully predict individual's DS, indicated by a significant correlation between the actual and predicted DS (r = 0.32, p < 0.01). Importantly, Prefrontal-Temporal, Insula-Limbic and Parietal-Occipital connections contributed to the prediction model. Furthermore, the model could also predict individual's daily distractibility in the other independent cohort (r = -0.34, p < 0.05). Our findings showed the efficiency of the predictive models of distractor suppression encompassing connections between large anatomical regions and highlighted the importance of the communications between attention-related and visual information processing regions in distractor suppression. Current findings may potentially provide neurobiological markers of visual distractor suppression.

Details DOI

EAAI Journal 2021 Journal Article

A multi-dimensional association information analysis approach to automated detection and localization of myocardial infarction

Jieshuo Zhang
Ming Liu
Peng Xiong
Haiman Du
Hong Zhang
Feng Lin
Zengguang Hou
Xiuling Liu

Developing an accurate and automatic algorithm for detection and localization of myocardial infarction (MI) remains a great challenge for multi-lead electrocardiograph (ECG) signals. The core is a novel technique of multi-dimensional association information analysis for a multi-lead ECG tensor. Tensorization based on Discrete Wavelet Transform is investigated to construct an effective ECG tensor containing multi-dimensional association information from 12-lead ECG signals. The multi-lead feature extraction algorithm based on Parallel Factor Analysis is developed to automatically extract the low-dimensional and highly recognizable lead characteristic features of the tensor. After that a bagged decision tree is constructed to categorize 12 types of heartbeats, healthy controls and 11 kinds of MI, from the lead features. Using the PTB database, we compare with the existing MI diagnosis methods. For MI detection, significant improvement of the accuracy, sensitivity and specificity are achieved; as high as 99. 88%, 99. 98% and 99. 39% respectively. Furthermore, an experiment with 36-dimensional features obtained from the ECG tensor is conducted for the localization of 11 kinds of MI, and our proposed method achieved an accuracy of 99. 40%, sensitivity of 99. 86%, and specificity of 99. 89%. The proposed algorithm can effectually accomplish the localization of 11 categories of MI by using the lead features extracted from the multi-dimensional association ECG tensor, which has not been achieved in literature. The accurate and comprehensive tool development will greatly help cardiologists diagnose 12-lead ECG signals of MI.

Details DOI

AAAI Conference 2021 Conference Paper

Auto-Encoding Transformations in Reparameterized Lie Groups for Unsupervised Learning

Feng Lin
Haohang Xu
Houqiang Li
Hongkai Xiong
Guo-Jun Qi

Unsupervised training of deep representations has demonstrated remarkable potentials in mitigating the prohibitive expenses on annotating labeled data recently. Among them is predicting transformations as a pretext task to self-train representations, which has shown great potentials for unsupervised learning. However, existing approaches in this category learn representations by either treating a discrete set of transformations as separate classes, or using the Euclidean distance as the metric to minimize the errors between transformations. None of them has been dedicated to revealing the vital role of the geometry of transformation groups in learning representations. Indeed, an image must continuously transform along the curved manifold of a transformation group rather than through a straight line in the forbidden ambient Euclidean space. This suggests the use of geodesic distance to minimize the errors between the estimated and groundtruth transformations. Particularly, we focus on homographies, a general group of planar transformations containing the Euclidean, similarity and affine transformations as its special cases. To avoid an explicit computing of intractable Riemannian logarithm, we project homographies onto an alternative group of rotation transformations SR(3) with a tractable form of geodesic distance. Experiments demonstrate the proposed approach to Auto-Encoding Transformations exhibits superior performances on a variety of recognition problems.

PDF Details

YNICL Journal 2020 Journal Article

Cognitively supernormal older adults maintain a unique structural connectome that is resistant to Alzheimer’s pathology

Quanjing Chen
Timothy M. Baran
Brian Rooks
M. Kerry O'Banion
Mark Mapstone
Zhengwu Zhang
Feng Lin

Studying older adults with excellent cognitive capacities (Supernormals) provides a unique opportunity for identifying factors related to cognitive success - a critical topic across lifespan. There is a limited understanding of Supernormals' neural substrates, especially whether any of them attends shaping and supporting superior cognitive function or confer resistance to age-related neurodegeneration such as Alzheimer's disease (AD). Here, applying a state-of-the-art diffusion imaging processing pipeline and finite mixture modelling, we longitudinally examine the structural connectome of Supernormals. We find a unique structural connectome, containing the connections between frontal, cingulate, parietal, temporal, and subcortical regions in the same hemisphere that remains stable over time in Supernormals, relatively to typical agers. The connectome significantly classifies positive vs. negative AD pathology at 72% accuracy in a new sample mixing Supernormals, typical agers, and AD risk [amnestic mild cognitive impairment (aMCI)] subjects. Among this connectome, the mean diffusivity of the connection between right isthmus cingulate cortex and right precuneus most robustly contributes to predicting AD pathology across samples. The mean diffusivity of this connection links negatively to global cognition in those Supernormals with positive AD pathology. But this relationship does not exist in typical agers or aMCI. Our data suggest the presence of a structural connectome supporting cognitive success. Cingulate to precuneus white matter integrity may be useful as a structural marker for monitoring neurodegeneration and may provide critical information for understanding how some older adults maintain or excel cognitively in light of significant AD pathology.

Details DOI

YNICL Journal 2019 Journal Article

How pattern information analyses of semantic brain activity elicited in language comprehension could contribute to the early identification of Alzheimer's Disease

Andrew James Anderson
Feng Lin

Alzheimer's disease (AD) is associated with a loss of semantic knowledge reflecting brain pathophysiology that begins years before dementia. Identifying early signs of pathophysiology induced dysfunction in the neural systems that access and process words' meaning could therefore help forecast dementia. This article reviews pioneering studies demonstrating that abnormal functional Magnetic Resonance Imaging (fMRI) response patterns elicited in semantic tasks reflect both AD-pathophysiology and the hereditary risk of AD, and also can help forecast cognitive decline. However, to bring current semantic task-based fMRI research up to date with new AD research guidelines the relationship with different types of AD-pathophysiology needs to be more thoroughly examined. We shall argue that new analytic techniques and experimental paradigms will be critical for this. Previous work has relied on specialized tests of specific components of semantic knowledge/processing (e.g. famous name recognition) to reveal coarse AD-related changes in activation across broad brain regions. Recent computational advances now enable more detailed tests of the semantic information that is represented within brain regions during more natural language comprehension. These new methods stand to more directly index how pathophysiology alters neural information processing, whilst using language comprehension as the basis for a more comprehensive examination of semantic brain function. We here connect the semantic pattern information analysis literature up with AD research to raise awareness to potential cross-disciplinary research opportunities.

Details DOI

EAAI Journal 2019 Journal Article

Multi-lead model-based ECG signal denoising by guided filter

Huaqing Hao
Ming Liu
Peng Xiong
Haiman Du
Hong Zhang
Feng Lin
Zengguang Hou
Xiuling Liu

The electrocardiogram (ECG) denoising is of paramount importance for accurate disease diagnosis, but individual differences bring great difficulties for ECG denoising, especially for Dynamic Electrocardiography (DCG). In this paper, a multi-lead model-based ECG signal denoising method is proposed, in which a guided filter is inherently adapted to denoise ECG signal. For each person, a patient-specific statistical model will be constructed by sparse autoencoder (SAE) which can effectively preserve the detailed signal features. Thus, the guided signal producing by the statistical model can perform well in the guided filter. Especially, even the sudden morphological changes, the denoised ECG signals can still be conserved. The results on the 12-lead Arrhythmia Database and the MIT-BIH Arrhythmia Database demonstrate that the signal-to-noise ratio (SNR) improvement of the proposed method can reach as high as 21. 54 dB, and the mean squared error (MSE) is less than 0. 0401. Besides achievement of minimum signal distortion in comparisons with the major of the current denoising algorithms for complex noise environment, the proposed method demonstrate robustness in the complex interferences, especially in tracing the sudden morphological changes of ECG signals. Due to the remarkable superiority in preserving diagnostic and detail features of ECG signals, the proposed method can handle ECG signals with abnormal heart beats, and then can improve the accuracy detection of the disease.

Details DOI

YNICL Journal 2018 Journal Article

Functional and structural connectivity of the amygdala underpins locus of control in mild cognitive impairment

Ping Ren
Benjamin Chapman
Zhengwu Zhang
Giovanni Schifitto
Feng Lin

Locus of control (LOC) is an important personality trait. LOC over cognitive competency reflects an individual's perceived control of desired cognitive outcomes, which is critical for maintaining successful cognitive aging. It is important to understand the neural substrates of LOC over cognitive competency in older adults, especially for individuals at high risk of dementia. Here, we characterized a cohesive functional and structural connectivity profile underlying LOC among 55 older adults with amnestic mild cognitive impairment (aMCI), combining resting-state functional magnetic resonance imaging and diffusion tensor imaging. The results showed that both functional and structural connectivity between the medial prefrontal cortex and amygdala were significantly correlated with external LOC. The functional connectivity mediated the correlation between structural connectivity and external LOC. In addition, aging-associated neurodegeneration moderated the relationship between structural connectivity and external LOC, showing that the structural connectivity was positively correlated with external LOC in low, but not high neurodegeneration. Our results suggest a critical role of the functional amygdala-frontal network, which may serve as a bridge between its white matter tract and LOC over cognitive competency in groups at high risk for dementia.

Details DOI

JBHI Journal 2018 Journal Article

Tempo-Spatial Compressed Sensing of Organ-on-a-Chip for Pervasive Health

Chen Song
Aosen Wang
Feng Lin
Mohammadnabi Asmani
Ruogang Zhao
Zhanpeng Jin
Jian Xiao
Wenyao Xu

As a micro-engineered biomimetic system to replicate key functions of living organs, organ-on-a-chip (OC) technology provides a high-throughput model for investigating complex cell interactions with both high temporal and spatial resolutions in biological studies. Typically, microscopy and high-speed video cameras are used for data acquisition, which are expensive and bulky. Recently, compressed sensing (CS) has increasingly attracted attentions due to its extremely low-complexity structure and low sampling rate. However, there is no CS solution tailored for tempo-spatial information acquisition. In this paper, we propose tempo-spatial CS (TS-CS), a unified CS architecture for OC stream, which achieves significant cost reduction and truly combines sensing with compression along the temporal and spatial domains. We point out that TS-CS can consistently achieve better performance by exploiting tempo-spatial compressibility in OC data. To this end, we comprehensively evaluate the system performance by employing four different bases for CS. With comparison to the traditional way, we show that TS-CS always obtains better recovery result with a throughput bound and can achieve around 25% throughput improvement under a reconstruction demand by applying discrete cosine transform matrix as the basis.

Details DOI

JBHI Journal 2017 Journal Article

Toward Unobtrusive Patient Handling Activity Recognition for Injury Reduction Among At-Risk Caregivers

Feng Lin
Aosen Wang
Lora Cavuoto
Wenyao Xu

Nurses regularly perform patient handling activities. These activities with awkward postures expose healthcare providers to a high risk of overexertion injury. The recognition of patient handling activities is the first step to reduce injury risk for caregivers. The current practice on workplace activity recognition is based on human observational approach, which is neither accurate nor projectable to a large population. In this paper, we aim at addressing these challenges. Our solution comprises a smart wearable device and a novel spatio-temporal warping (STW) pattern recognition framework. The wearable device, named Smart Insole 2. 0, is equipped with a rich set of sensors and can provide an unobtrusive way to automatically capture the information of patient handling activities. The STW pattern recognition framework fully exploits the spatial and temporal characteristics of plantar pressure by calculating a novel warped spatio-temporal distance, to quantify the similarity for the purpose of activity recognition. To validate the effectiveness of our framework, we perform a pilot study with eight subjects, including eight common activities in a nursing room. The experimental results show the overall classification accuracy achieves 91. 7%. Meanwhile, the qualitative profile and load level can also be classified with accuracies of 98. 3% and 92. 5%, respectively.

Details DOI

ICRA Conference 2001 Conference Paper

An Optimal Deadlock Avoidance Policy for Manufacturing System with Flexible Operation Sequence and Flexible Routing

Keyi Xing
Feng Lin
Baosheng Hu

Addresses the deadlock avoidance problems in manufacturing systems with sequence and route flexibility. It is shown that for a large class of such systems, through one-step look ahead, an optimal deadlock avoidance policy can be obtained. For a general system the authors present a system reduction method by which the system is reduced and the reduced system has an optimal deadlock avoidance policy. Using this optimal control policy and restricting the allocation of resources with capacity one, a sub-optimal deadlock avoidance policy for the system is obtained. The proposed policies have polynomial computation complexity.

Details

ICRA Conference 1999 Conference Paper

Discrete Event Control with Active Events

Michael Heymann
Feng Lin
George Meyer
Satya Ranjan Mohanty
Vigyan Chandra
Ratnesh Kumar 0001

The traditional framework for discrete-event control is extended to include the case of control with active events, in which both the user and the environment have events that they can trigger. A variety of liveness and safety specifications can be considered within this extended framework. A synthesis algorithm of minimally restrictive controllers is outlined.

Details