Arrow Research search

Author name cluster

Min Xu

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.

39 papers
2 author rows

Possible papers

39

AAAI Conference 2026 Conference Paper

DiLO: Disentangled Latent Optimization for Learning Shape and Deformation in Grouped Deforming 3D Objects

  • Mostofa Rafid Uddin
  • Jana Armouti
  • Umong Sain
  • Md Asib Rahman
  • Xingjian Li
  • Min Xu

In this work, we propose a disentangled latent optimization-based method for parameterizing grouped deforming 3D objects into shape and deformation factors in an unsupervised manner. Our approach involves the joint optimization of a generator network along with the shape and deformation factors, supported by specific regularization techniques. For efficient amortized inference of disentangled shape and deformation codes, we train two order-invariant PoinNet-based encoder networks in the second stage of our method. We demonstrate several significant downstream applications of our method, including unsupervised deformation transfer, deformation classification, and explainability analysis. Extensive experiments conducted on 3D human, animal, and facial expression datasets demonstrate that our simple approach is highly effective in these downstream tasks, comparable or superior to existing methods with much higher complexity.

PDF Details DOI

TMLR Journal 2026 Journal Article

Prompt-based Adaptation in Large-scale Vision Models: A Survey

  • Xi Xiao
  • Yunbei Zhang
  • Lin Zhao
  • Yiyang Liu
  • Xiaoying Liao
  • Zheda Mai
  • Xingjian Li
  • Xiao Wang

In computer vision, Visual Prompting (VP) and Visual Prompt Tuning (VPT) have recently emerged as lightweight and effective alternatives to full fine-tuning for adapting large-scale vision models within the ``pretrain-then-finetune'' paradigm. However, despite rapid progress, their conceptual boundaries remain blurred, as VP and VPT are frequently used interchangeably in current research, reflecting a lack of systematic distinction between these techniques and their respective applications. In this survey, we revisit the designs of VP and VPT from first principles, and conceptualize them within a unified framework termed Prompt-based Adaptation (PA). Within this framework, we distinguish methods based on their injection granularity: VP operates at the pixel level, while VPT injects prompts at the token level. We further categorize these methods by their generation mechanism into fixed, learnable, and generated prompts. Beyond the core methodologies, we examine PA’s integrations across diverse domains, including medical imaging, 3D point clouds, and vision-language tasks, as well as its role in test-time adaptation and trustworthy AI. We also summarize current benchmarks and identify key challenges and future directions. To the best of our knowledge, we are the first comprehensive survey dedicated to PA's methodologies and applications in light of their distinct characteristics. Our survey aims to provide a clear roadmap for researchers and practitioners in all area to understand and explore the evolving landscape of PA-related research.

PDF Details

YNIMG Journal 2025 Journal Article

Developmental changes in brain activation and functional connectivity during Chinese handwriting

  • Huan Ren
  • Junjun Li
  • Zhenglong Lin
  • Yanzhi Xu
  • Hong-Yan Bi
  • Min Xu
  • Yang Yang

Handwriting is a fundamental linguistic and motor skill that significantly contributes to cognitive development, effective communication, and academic success. However, how brain systems develop to support handwriting remains unclear. Using functional magnetic resonance imaging (fMRI), this study examined the differences in regional activation and functional network connectivity during Chinese handwriting between 52 children (9-13 years old) and 40 young adults. We found that, compared to adults, children exhibited lower activation in the left inferior frontal gyrus, left inferior parietal lobule, left precentral gyrus, and right posterior occipital cortex, possibly reflecting variations in functional specialization of handwriting-related circuits. In contrast, network analysis showed that children had greater global and local efficiency of the whole-brain networks than adults, particularly in the default mode network and visual network, possibly reflecting a large-scale reconfiguration of brain networks during handwriting automatization. These observed differences suggest that the brain systems supporting handwriting are still being established from middle childhood to young adulthood, underpinned by focal functional specialization and reconfiguration of functional networks. These findings provide new insights into the neurodevelopment of skills that integrate linguistic and motor processes, and expand our understanding of brain functional plasticity.

Details DOI

TMLR Journal 2025 Journal Article

Genetic-Evolutionary Graph Neural Networks: A Paradigm for Improved Graph Representation Learning

  • Haimin Zhang
  • Min Xu

Message-passing graph neural networks have become the dominant framework for learning over graphs. However, empirical studies continually show that message-passing graph neural networks tend to generate over-smoothed representations for nodes after iteratively applying message passing. This over-smoothing problem is a core issue that limits the representational capacity of message-passing graph neural networks. We argue that the fundamental problem with over-smoothing is a lack of diversity in the generated embeddings, and the problem could be reduced by enhancing the embedding diversity in the embedding generation process. To this end, we propose genetic-evolutionary graph neural networks, a new paradigm for graph representation learning inspired by genetic algorithms. We view each layer of a graph neural network as an evolutionary process and develop operations based on crossover and mutation to prevent embeddings from becoming similar to one another, thus enabling the model to generate improved graph representations. The proposed framework has good interpretablility, as it directly draws inspiration from genetic algorithms for preserving population diversity. We experimentally validate the proposed framework on six benchmark datasets on different tasks. The results show that our method significant advances the performance of current graph neural networks, resulting in new state-of-the-art results for graph representation learning on these datasets.

PDF Details

IROS Conference 2025 Conference Paper

Improved Calibration for Panoramic Annular Lens Systems with Angular Modulation

  • Ding Wang
  • Junhua Wang
  • Yuhan Tian
  • Min Xu
  • Lingbao Kong

This paper addresses the challenges of calibrating Panoramic Annular Lens (PAL) systems, which exhibit unique projection characteristics due to their imaging relationship designed to compress blind zones. Traditional camera calibration methods often fail to accurately capture these properties. To resolve this limitation, we propose a novel projection model that incorporates angular modulation, enabling a more accurate representation of the PAL system’s imaging process. This formulation significantly improves the model’s ability to describe the relationship between object space and image space. We evaluate our approach on both synthetic and real-world datasets tailored for PAL cameras. Experimental results demonstrate that the model achieves sub-pixel accuracy, with reprojection errors typically ranging from 0. 1 to 0. 3 pixels on 2048×2048 images when using five distortion terms. This level of precision surpasses existing calibration models for panoramic cameras, making our method particularly suitable for high-accuracy applications. The datasets used in this study are publicly available at https://github.com/wwendy233/PALcalib.

Details

JBHI Journal 2025 Journal Article

OASIS-Net: An Obstetric Adversarial Semi-Supervised Image Segmentation Network for Cervical and Fetal Head Ultrasound Imaging

  • Minh Huu Nhat Le
  • Tran Quoc Khanh Le
  • Thanh-Huy Nguyen
  • Dang Nguyen
  • Hien Quang Nguyen
  • Nhi Huu Hanh Le
  • Kien Dang Nguyen
  • Hien Quang Kha

Accurate obstetric ultrasound segmentation is hampered by speckle noise and scarce annotations. We propose OASIS-Net, a dual-space adversarial semi-supervised framework that trains a single DeepLabV3 $+$ backbone by minimizing one unified consistency loss. The loss couples input-space adversaries (iterative FGSM with $K=3$ steps, $\epsilon =4/255$ ) and weight-space gradient-aligned perturbations (DGAP, weight scale $=0. 5$ ) whose influence grows with a sigmoid ramp ( $T_{\text{ramp}}=20$, $\alpha _{\max }=1. 0$ ). Pseudo-labels are accepted with a confidence threshold of 0. 95 and the unlabeled loss weight is 1. 0. We evaluate OASIS-Net on two public obstetric benchmarks: FUGC (50 labeled, 450 unlabeled) and PSFH (5, 101 frames, 70% unlabeled). Using 20% of labels, the method attains Dice = 96. 53% and HD $_{95}$ = 3. 86 px on FUGC, and Dice = 97. 16% and HD $_{95}$ = 2. 34 px on PSFH. Ablation shows that removing either perturbation stream reduces Dice by up to 1. 8 percentage points. The trained model runs at 18. 96 frames s $^{-1}$ on a single RTX 4060 Ti and produces high-precision masks that enable automated cervical-length and angle-of-progression measurements for objective obstetric screening and intrapartum monitoring. These results demonstrate that jointly enforcing input- and parameter-space adversarial consistency yields a label-efficient, robust solution for obstetric ultrasound segmentation and supports real-time clinical use

Details DOI

AAAI Conference 2025 Conference Paper

OLMD: Orientation-aware Long-term Motion Decoupling for Continuous Sign Language Recognition

  • Yiheng Yu
  • Sheng Liu
  • Yuan Feng
  • Min Xu
  • Zhelun Jin
  • Xuhua Yang

The primary challenge in continuous sign language recognition (CSLR) mainly stems from the presence of multi-orientational and long-term motions. However, current research overlooks these crucial aspects, significantly impacting accuracy. To tackle these issues, we propose a novel CSLR framework: Orientation-aware Long-term Motion Decoupling (OLMD), which efficiently aggregates long-term motions and decouples multi-orientational signals into easily interpretable components. Specifically, our innovative Long-term Motion Aggregation (LMA) module filters out static redundancy while adaptively capturing abundant features of long-term motions. We further enhance orientation awareness by decoupling complex movements into horizontal and vertical components, allowing for motion purification in both orientations. Additionally, two coupling mechanisms are proposed: stage and cross-stage coupling, which together enrich multi-scale features and improve the generalization capabilities of the model. Experimentally, OLMD shows SOTA performance on three large-scale datasets: PHOENIX14, PHOENIX14-T, and CSL-Daily. Notably, we improve the word error rate (WER) on PHOENIX14 by an absolute 1.6% compared to the previous SOTA.

PDF Details DOI

YNIMG Journal 2025 Journal Article

Syntactic development and its interplay with word processing and working memory in preschoolers’ brain: An fNIRS longitudinal study

  • Dongsu Yan
  • Tongfu Fang
  • Wei He
  • Min Xu

Understanding how children acquire syntactic structures from a limited set of grammatical rules and use them creatively to convey meaning has been a longstanding interest for scientific communities. Previous studies on syntactic development have revealed its close correlation with the development of vocabulary and working memory. Our study sought to elucidate how the relations between syntactic processing, word processing, and working memory were instantiated in the brain, and how earlier neural patterns might predict language abilities one year later. We employed functional near-infrared spectroscopy to examine among preschool children (N=50, Mage=61. 5 months) the neural activation associated with processing sentences of varying syntactic complexities, as well as tasks assessing word comprehension and working memory. The results revealed greater brain activation in the left inferior frontal gyrus (IFG) for syntactically complex as compared to simple sentences, and the activation magnitude was correlated with working memory. There was also a link between neural activity for sentence comprehension and word comprehension in bilateral superior temporal regions (STG). Moreover, the inter-regional and inter-hemispheric connectivity of IFG and STG/MTG could successfully predict children's future language comprehension one year later. The findings provide new insights into how the brain supports syntactic development and its interplay with word processing and working memory.

Details DOI

AAAI Conference 2025 Conference Paper

Vox-UDA: Voxel-wise Unsupervised Domain Adaptation for Cryo-Electron Subtomogram Segmentation with Denoised Pseudo-Labeling

  • Haoran Li
  • Xingjian Li
  • Jiahua Shi
  • Huaming Chen
  • Bo Du
  • Daisuke Kihara
  • Johan Barthelemy
  • Jun Shen

Cryo-Electron Tomography (cryo-ET) is a 3D imaging technology that facilitates the study of macromolecular structures at near-atomic resolution. Recent volumetric segmentation approaches on cryo-ET images have drawn widespread interest in the biological sector. However, existing methods heavily rely on manually labeled data, which requires highly professional skills, thereby hindering the adoption of fully-supervised approaches for cryo-ET images. Some unsupervised domain adaptation (UDA) approaches have been designed to enhance the segmentation network performance using unlabeled data. However, applying these methods directly to cryo-ET image segmentation tasks remains challenging due to two main issues: 1) the source dataset, usually obtained through simulation, contains a fixed level of noise, while the target dataset, directly collected from raw-data from the real-world scenario, have unpredictable noise levels. 2) the source data used for training typically consists of known macromoleculars. In contrast, the target domain data are often unknown, causing the model to be biased towards those known macromolecules, leading to a domain shift problem. To address such challenges, in this work, we introduce a voxel-wise unsupervised domain adaptation approach, termed Vox-UDA, specifically for cryo-ET subtomogram segmentation. Vox-UDA incorporates a noise generation module to simulate target-like noises in the source dataset for cross-noise level adaptation. Additionally, we propose a denoised pseudo-labeling strategy based on the improved Bilateral Filter to alleviate the domain shift problem. More importantly, we construct the first UDA cryo-ET subtomogram segmentation benchmark on three experimental datasets. Extensive experimental results on multiple benchmarks and newly curated real-world datasets demonstrate the superiority of our proposed approach compared to state-of-the-art UDA methods.

PDF Details DOI

AAAI Conference 2024 Conference Paper

Deep Active Learning with Noise Stability

  • Xingjian Li
  • Pengkun Yang
  • Yangcheng Gu
  • Xueying Zhan
  • Tianyang Wang
  • Min Xu
  • Chengzhong Xu

Uncertainty estimation for unlabeled data is crucial to active learning. With a deep neural network employed as the backbone model, the data selection process is highly challenging due to the potential over-confidence of the model inference. Existing methods resort to special learning fashions (e.g. adversarial) or auxiliary models to address this challenge. This tends to result in complex and inefficient pipelines, which would render the methods impractical. In this work, we propose a novel algorithm that leverages noise stability to estimate data uncertainty. The key idea is to measure the output derivation from the original observation when the model parameters are randomly perturbed by noise. We provide theoretical analyses by leveraging the small Gaussian noise theory and demonstrate that our method favors a subset with large and diverse gradients. Our method is generally applicable in various tasks, including computer vision, natural language processing, and structural data analysis. It achieves competitive performance compared against state-of-the-art active learning baselines.

PDF Details DOI

YNIMG Journal 2024 Journal Article

Differential neural representations of syntactic and semantic information across languages in Chinese-English bilinguals

  • Zeqi Hou
  • Hehui Li
  • Lin Gao
  • Jian Ou
  • Min Xu

Bilingual individuals manage multiple languages that align in conceptual meaning but differ in forms and structures. While prior research has established foundational insights into the neural mechanisms in bilingual processing, the extent to which the first (L1) and second language (L2) systems overlap or diverge across different linguistic components remains unclear. This study probed the neural underpinnings of syntactic and semantic processing for L1 and L2 in Chinese-English bilinguals (N = 44) who performed sentence comprehension tasks and an N-back working memory task during functional MRI scanning. We observed that the increased activation for L2 processing was within the verbal working memory network, suggesting a greater cognitive demand for processing L2. Crucially, we looked for brain regions showing adaptation to the repetition of semantic information and syntactic structure, and found more robust adaptation effects in L1 in the middle and superior temporal cortical areas. The differential adaptation effects between L1 and L2 were more pronounced for the semantic condition. Multivariate pattern analysis further revealed distinct neural sensitivities to syntactic and semantic representations between L1 and L2 across frontotemporal language regions. Our findings suggest that while L1 and L2 engage similar neural systems, finer representation analyses uncover distinct neural patterns for both semantic and syntactic aspects in the two languages. This study advances our understanding of neural representations involved in different language components in bilingual individuals.

Details DOI

AAAI Conference 2024 Conference Paper

MedSegDiff-V2: Diffusion-Based Medical Image Segmentation with Transformer

  • Junde Wu
  • Wei Ji
  • Huazhu Fu
  • Min Xu
  • Yueming Jin
  • Yanwu Xu

The Diffusion Probabilistic Model (DPM) has recently gained popularity in the field of computer vision, thanks to its image generation applications, such as Imagen, Latent Diffusion Models, and Stable Diffusion, which have demonstrated impressive capabilities and sparked much discussion within the community. Recent investigations have further unveiled the utility of DPM in the domain of medical image analysis, as underscored by the commendable performance exhibited by the medical image segmentation model across various tasks. Although these models were originally underpinned by a UNet architecture, there exists a potential avenue for enhancing their performance through the integration of vision transformer mechanisms. However, we discovered that simply combining these two models resulted in subpar performance. To effectively integrate these two cutting-edge techniques for the Medical image segmentation, we propose a novel Transformer-based Diffusion framework, called MedSegDiff-V2. We verify its effectiveness on 20 medical image segmentation tasks with different image modalities. Through comprehensive evaluation, our approach demonstrates superiority over prior state-of-the-art (SOTA) methodologies. Code is released at https://github.com/KidsWithTokens/MedSegDiff.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

Metric from Human: Zero-shot Monocular Metric Depth Estimation via Test-time Adaptation

  • Yizhou Zhao
  • Hengwei Bian
  • Kaihua Chen
  • Pengliang Ji
  • Liao Qu
  • Shao-yu Lin
  • Weichen Yu
  • Haoran Li

Monocular depth estimation (MDE) is fundamental for deriving 3D scene structures from 2D images. While state-of-the-art monocular relative depth estimation (MRDE) excels in estimating relative depths for in-the-wild images, current monocular metric depth estimation (MMDE) approaches still face challenges in handling unseen scenes. Since MMDE can be viewed as the composition of MRDE and metric scale recovery, we attribute this difficulty to scene dependency, where MMDE models rely on scenes observed during supervised training for predicting scene scales during inference. To address this issue, we propose to use humans as landmarks for distilling scene-independent metric scale priors from generative painting models. Our approach, Metric from Human (MfH), bridges from generalizable MRDE to zero-shot MMDE in a generate-and-estimate manner. Specifically, MfH generates humans on the input image with generative painting and estimates human dimensions with an off-the-shelf human mesh recovery (HMR) model. Based on MRDE predictions, it propagates the metric information from painted humans to the contexts, resulting in metric depth estimations for the original input. Through this annotation-free test-time adaptation, MfH achieves superior zero-shot performance in MMDE, demonstrating its strong generalization ability.

PDF Details DOI

TCS Journal 2023 Journal Article

Relationship between diagnosability and non-inclusive diagnosability of triangle-free connected graphs under the PMC model

  • Tongtong Ding
  • Min Xu

Diagnosability, defined as the maximum number of fault processors that the system can recognize, is an important parameter in the design and maintenance of multiprocessor systems. To improve diagnosability, Ding et al. [5] proposed the non-inclusive diagnosability. In this paper, we discuss the relationship between the diagnosability and non-inclusive diagnosability of triangle-free connected graphs under the PMC model. We use the non-inclusive diagnosability of some regular graphs as examples in our discussion.

Details DOI

TCS Journal 2023 Journal Article

Two-disjoint-cycle-cover vertex bipancyclicity of bipartite hypercube-like networks

  • Ruichao Niu
  • Shujie Zhou
  • Min Xu

Let r 1, r 2 be two integers such that r 2 ≥ r 1 ≥ 0. A bipartite graph G is two-disjoint-cycle-cover vertex [ r 1, r 2 ] -bipancyclic (2-DCC vertex [ r 1, r 2 ] -bipancyclic for short) if for any two vertices u, v ∈ V ( G ) and any even integer ℓ satisfying r 1 ≤ ℓ ≤ r 2, there exist two vertex-disjoint cycles C 1 and C 2 in G with | V ( C 1 ) | = ℓ and | V ( C 2 ) | = | V ( G ) | − ℓ such that u ∈ V ( C 1 ) and v ∈ V ( C 2 ). In this paper, we study the 2-DCC vertex bipancyclicity of the n-dimensional bipartite hypercube-like network, which is one class of hypercube-generalized networks. As a consequence, we show that an n-dimensional bipartite hypercube-like network is 2-DCC vertex [ 4, 2 n − 1 ] -bipancyclic for n ≥ 3. In particular, it provides an application that n-dimensional hypercube and bicube are also 2-DCC vertex [ 4, 2 n − 1 ] -bipancyclic for n ≥ 3.

Details DOI

AAAI Conference 2022 Conference Paper

Boosting Active Learning via Improving Test Performance

  • Tianyang Wang
  • Xingjian Li
  • Pengkun Yang
  • Guosheng Hu
  • Xiangrui Zeng
  • Siyu Huang
  • Cheng-Zhong Xu
  • Min Xu

Central to active learning (AL) is what data should be selected for annotation. Existing works attempt to select highly uncertain or informative data for annotation. Nevertheless, it remains unclear how selected data impacts the test performance of the task model used in AL. In this work, we explore such an impact by theoretically proving that selecting unlabeled data of higher gradient norm leads to a lower upper-bound of test loss, resulting in a better test performance. However, due to the lack of label information, directly computing gradient norm for unlabeled data is infeasible. To address this challenge, we propose two schemes, namely expected-gradnorm and entropy-gradnorm. The former computes the gradient norm by constructing an expected empirical loss while the latter constructs an unsupervised loss with entropy. Furthermore, we integrate the two schemes in a universal AL framework. We evaluate our method on classical image classification and semantic segmentation tasks. To demonstrate its competency in domain applications and its robustness to noise, we also validate our method on a cellular imaging analysis task, namely cryo-Electron Tomography subtomogram classification. Results demonstrate that our method achieves superior performance against the state of the art. We refer readers to https: //arxiv. org/pdf/2112. 05683. pdf for the full version of this paper which includes the appendix and source code link.

PDF Details

YNICL Journal 2022 Journal Article

Radiomics for prediction of intracerebral hemorrhage outcomes: A retrospective multicenter study

  • Xiaoyu Huang
  • Dan Wang
  • Qiaoying Zhang
  • Yaqiong Ma
  • Hui Zhao
  • Shenglin Li
  • Juan Deng
  • Jialiang Ren

BACKGROUND: Accurate risk stratification of patients with intracerebral hemorrhage (ICH) could help refine adjuvant therapy selection and better understand the clinical course. We aimed to evaluate the value of radiomics features from hematomal and perihematomal edema areas for prognosis prediction and to develop a model combining clinical and radiomic features for accurate outcome prediction of patients with ICH. METHODS: This multicenter study enrolled patients with ICH from January 2016 to November 2021. Their outcomes at 3 months were recorded based on the modified Rankin Scale (good, 0-3; poor, 4-6). Independent clinical and radiomic risk factors for poor outcome were identified through multivariate logistic regression analysis, and predictive models were developed. Model performance and clinical utility were evaluated in both internal and external cohorts. RESULTS: Among the 1098 ICH patients evaluated (mean age, 60 ± 13 years), 703 (64 %) had poor outcomes. Age, hemorrhage volume and location, and Glasgow Coma Scale (GCS) were independently associated with outcomes. The area under the receiver operating characteristic curve (AUC) of the clinical model was 0.881 in the external validation cohort. Addition of the Rad-score (combined hematoma and perihematomal edema area) improved predictive accuracy and model performance (AUC, 0.893), net reclassification improvement, 0.140 (P < 0.001), and integrated discrimination improvement, 0.050 (P < 0.001). CONCLUSIONS: The radiomics features of hematomal and perihematomal edema area have additional value in prognostic prediction; moreover, addition of radiomic features significantly improves model accuracy.

Details DOI

TCS Journal 2022 Journal Article

The unpaired many-to-many k-disjoint paths in bipartite hypercube-like networks

  • Ruichao Niu
  • Min Xu

In this paper, we discuss about the problem of unpaired many-to-many disjoint paths in bipartite hypercube-like networks with faulty elements. Let F be a faulty set with f faulty elements in an n-dimensional bipartite hypercube-like network G. For every positive integer k ≥ 1 satisfying f + k ≤ n − 1, and two arbitrary sets S and T with k fault-free vertices in distinct parts of the bipartition of G, there exists an unpaired fault-free k-disjoint path collection joining S and T in G − F, which contains at least 2 n − 2 f v vertices, where f v is the number of faulty vertices in G. This result is optimal in the worst case, and the bound f + k ≤ n − 1 is tight.

Details DOI

TCS Journal 2021 Journal Article

Conditional diagnosability of Cayley graphs generated by wheel graphs under the PMC model

  • Yulong Wei
  • Min Xu

Fault diagnosis of systems is an important area of study in the design and maintenance of multiprocessor systems. In 2005, Lai et al. [12] introduced conditional diagnosability under the assumption that all the neighbors of any processor in a multiprocessor system cannot be faulty at the same time. In this paper, we completely determine the conditional diagnosability of Cayley graphs generated by wheel graphs W G n under the PMC model.

Details DOI

TCS Journal 2021 Journal Article

Symmetric PMC model of diagnosis, b-matchings in graphs and fault identification in t-diagnosable systems

  • Qiang Zhu
  • Krishnaiyan Thulasiraman
  • Kshirasagar Naik
  • Sridhar Radhakrishnan
  • Min Xu

A major breakthrough in the diagnosis of t-diagnosable systems occurred when Dahbura and Mason developed a diagnosis algorithm under the PMC model using the matching theory in bipartite graphs. In this paper we introduce a new testing model called the symmetric PMC(SPMC) model. Our main contributions are as follows: We prove that the diagnosability of an n-dimensional hypercube under the SPMC model is almost twice its diagnosability under the PMC model. We then show that the fault diagnosis problem for a t-diagnosable system under the SPMC model reduces to that of determining a maximum weighted b-matching in its diagnosis graph. Algorithm LABEL is then given to identify all the faulty vertices using the maximum b-matching of the diagnosis graph. Finally, using certain results from the theory of partitions of integers we establish the worst case complexity of our t-diagnosis algorithm. The complexity is much better than the worst case complexity of the Dahbura-Mason algorithm for the t-diagnosable problem under the PMC model.

Details DOI

TCS Journal 2020 Journal Article

Edge-fault-tolerant strong Menger edge connectivity on regular graphs

  • Min Xu
  • Pingshan Li

The connectivity of a graph is an important issue in graph theory and is also one of the most important factors in evaluating the reliability and fault tolerance of a network. A graph G is called m-edge-fault-tolerant strongly Menger (m-EFTSM for short) edge connected if there are min ⁡ { deg G − F ⁡ ( x ), deg G − F ⁡ ( y ) } edge-disjoint paths between any two different vertices x and y in G − F for any F ⊆ E ( G ) with | F | ≤ m. In this paper, we give a necessary and sufficient condition of EFTSM edge connectivity on regular graphs. And we obtain several optimal results about EFTSM edge connectivity on (1, 2)-matching composition networks, each of which is constructed by connecting two graphs via one or two perfect matchings. As applications, we show that the class of n-dimensional hypercube-like networks (included hypercube, crossed cube et al.) are ( n − 2 ) -EFTSM edge connected; show that the n-dimensional folded hypercube is ( n − 1 ) -EFTSM edge connected, and show that the n-dimensional augmented cube is ( 2 n − 3 ) -EFTSM edge connected. The bounds ( n − 2 ), ( n − 1 ) and ( 2 n − 3 ) are sharp.

Details DOI

TCS Journal 2020 Journal Article

Fault tolerance of hypercube like networks: Spanning laceability under edge faults

  • Min Xu
  • Kshirasagar Naik
  • Krishnaiyan Thulasiraman

Given two vertices u and v in a connected undirected graph G, a w⁎-container C ( u, v ) is a set of w internally vertex disjoint paths between u and v spanning all the vertices in G. A bipartite graph G is w ⁎ -laceable if there exists a w ⁎ -container between any two vertices belonging to different partitions of G. In [8], [33] a class B n ′ of bipartite graphs called hypercube-like bipartite networks was defined. In [22], Lin et al. showed that every graph in B n ′ is w ⁎ -laceable for every 1 ≤ w ≤ n. We define a graph is f-edge fault tolerant w ⁎ -laceable if G − F is w ⁎ -laceable for any arbitrary subset F of edges of G with | F | ≤ f. In this paper we show that every graph in B n ′ is f-edge-fault tolerant w ⁎ -laceable for every 0 ≤ f ≤ n − 2 and 1 ≤ w ≤ n − f which generalize Lin's result. We also give generalization of two other results in [22, 27].

Details DOI

TCS Journal 2020 Journal Article

The component (edge) connectivity of shuffle-cubes

  • Tongtong Ding
  • Pingshan Li
  • Min Xu

Component (edge) connectivity is a generalization of traditional (edge) connectivity. Let F be a vertex set (resp. , an edge set), if G − F has at least g components, F is a g-component (edge) cut. The g-component (edge) connectivity of graph G is the minimum size of the g-component (edge) cut. In this paper, we study the g-component (edge) connectivity of shuffle-cubes for small g.

Details DOI

TCS Journal 2020 Journal Article

The largest component of faulty star graphs

  • Pingshan Li
  • Min Xu

The minimum size of the largest component of a network with faults is a useful parameter to make a full evaluation of this network and it is helpful for estimating some other parameters in graph theory. In this paper, we focus on discussing the minimum size of the largest component of star graph S n which contains faulty edges. Specifically, we show that there exists at most one, two, three, four, five vertices beyond the largest component of S n − F with F ⊆ E ( S n ) and | F | ≤ 2 n − 5, 3 n − 8, 4 n − 11, 5 n − 15, 6 n − 19 respectively. As applications, we study the g-extra edge connectivity of S n for 1 ≤ g ≤ 5, and study the strong Menger edge connectivity of S n − F with F ∈ E ( S n ).

Details DOI

TCS Journal 2019 Journal Article

Hybrid fault diagnosis capability analysis of regular graphs

  • Yulong Wei
  • Min Xu

Fault diagnosis of systems is an important area of study in the design and maintenance of multiprocessor systems. A new measure for fault diagnosis of systems, namely, h-edge tolerable diagnosability, is the maximum number of faulty nodes that a system G can guarantee to locate when the number of faulty links does not exceed h, denoted by t h e ( G ). Particularly, t 0 e ( G ) is the traditional diagnosability of G. Chang et al. (2005) [5] determined the traditional diagnosabilities of regular graphs. In this paper, we establish the h-edge tolerable diagnosabilities of regular graphs under the PMC model and the MM⁎ model, which extends their results and provides a more precise characterization for the fault diagnosis capability of regular graphs. As applications, the h-edge tolerable diagnosabilities of many networks are determined under the PMC model and the MM⁎ model.

Details DOI

TCS Journal 2019 Journal Article

The h-edge tolerable diagnosability of balanced hypercubes

  • Min Xu
  • Yulong Wei

To measure the fault diagnosis capability of a multiprocessor system with faulty links, Zhu et al. (2019) [26] introduced the h-edge tolerable diagnosability. This kind of diagnosability is a generalization of the concept of traditional diagnosability. In this paper, as complement to the results in Wei and Xu (2019) [16], we completely determine the h-edge tolerable diagnosability of balanced hypercubes B H n under the PMC model and the MM⁎ model. Thus, the traditional diagnosability of B H n is also determined.

Details DOI

TCS Journal 2019 Journal Article

The t/k-diagnosability and strong Menger connectivity on star graphs with conditional faults

  • Pingshan Li
  • Min Xu

The t / k -diagnosis, a famous diagnosis strategy, is proposed by Somani and Peleg. In this strategy, all the faulty vertices, no more than t, can be isolated into a faulty set, which may contain no more than k fault-free vertices. Somani and Peleg ([16], 1996) first proved that S n was [ ( k + 1 ) n − 3 k − 2 ] / k -diagnosable for 0 < k ≤ n. Chen and Liu ([4], 2012) found that the proof of that proposition was flawed and they pointed out that S n was actually [ ( k + 1 ) n − 3 k − 1 ] / k -diagnosable for small k without a specific rang of k. Zhou et al. ([21], 2015) proved that S n ( n ≥ 5 ) was [ ( k + 1 ) n − 3 k − 1 ] / k -diagnosable for 1 ≤ k ≤ 3 and they proposed a problem: was S n still [ ( k + 1 ) n − 3 k − 1 ] / k -diagnosable for k ≥ 4? In this paper, we prove that S n ( n ≥ 5 ) is ( 5 n − 14 ) / 4 -diagnosable and it is not ( 5 n − 13 ) / 4 -diagnosable, which gives a negative answer to the problem of Zhou et al. Furthermore, we show that S n is still strongly Menger connected even if ( 4 n − 13 ) vertices fail, which means there are min ⁡ { deg G − F ⁡ ( x ), deg G − F ⁡ ( y ) } -internally vertex disjoint paths between any two different vertices x, y in G − F for F ∈ V ( S n ), δ ( G − F ) ≥ 2 and | F | ≤ 4 n − 13, and the bound 4 n − 13 is sharp.

Details DOI

TCS Journal 2018 Journal Article

Conditional (edge-)fault-tolerant strong Menger (edge) connectivity of folded hypercubes

  • Qi Cheng
  • Pingshan Li
  • Min Xu

Menger's theorem is a characterization of the connectivity in finite graphs in terms of the minimum number of disjoint paths that can be found between any pair of vertices. According to Menger's theorem, a graph G is k-connected if and only if any two vertices of G are connected by at least k internally disjoint paths. Moreover, there are at least κ ( G ) internally disjoint paths and, at most, min ⁡ { deg G ⁡ ( u ), deg G ⁡ ( v ) } internally disjoint paths between any two distinct vertices u, v in G. Motivated by this observation, Oh and Chen (resp. , Qiao and Yang) proposed the (fault-tolerant) strong Menger (resp. , edge) connectivity as follows. A connected graph G is called strongly Menger (edge) connected if for any two distinct vertices x, y in G, there are min ⁡ { deg G ⁡ ( x ), deg G ⁡ ( y ) } (edge-)disjoint paths between x and y. A graph G is called m-(edge-)fault-tolerant strongly Menger (edge) connected if G − F remains strongly Menger (edge) connected for an arbitrary set F ⊆ V ( G ) (resp. , F ⊆ E ( G ) ) with | F | ≤ m. A graph G is called m-conditional (edge-)fault-tolerant strongly Menger (edge) connected if G − F remains strongly Menger (edge) connected for an arbitrary set F ⊆ V ( G ) (resp. , F ⊆ E ( G ) ), | F | ≤ m and δ ( G − F ) ≥ 2. Qiao and Yang (2017) proved that all n-dimensional folded hypercubes are ( 2 n − 2 ) -conditional edge-fault-tolerant strongly Menger edge connected for n ≥ 5. Yang, Zhao and Zhang (2017) showed that all n-dimensional folded hypercubes are ( 2 n − 3 ) -conditional fault-tolerant strongly Menger connected for n ≥ 8. In this paper, we improve the result of Qiao and Yang by showing that all n-dimensional folded hypercubes are ( 3 n − 5 ) -conditional edge-fault-tolerant strongly Menger edge connected for n ≥ 5. Moreover, we present an example to show that our result is optimal with respect to the maximum tolerated edge faults. In addition, we show that the result of Yang, Zhao and Zhang is optimal by proving that the n-dimensional folded hypercubes are not ( 2 n − 2 ) -conditional fault-tolerant strongly Menger connected for n ≥ 8.

Details DOI

TCS Journal 2018 Journal Article

Fault-tolerant strong Menger (edge) connectivity and 3-extra edge-connectivity of balanced hypercubes

  • Pingshan Li
  • Min Xu

A connected graph G is called strongly Menger (edge) connected if for any two distinct vertices x, y of G, there are min ⁡ { deg G ⁡ ( x ), deg G ⁡ ( y ) } (-edge)-disjoint paths between x and y. The g ( g ≥ 0 ) -extra edge-connectivity of the connected graph G, denoted by λ g ( G ), is the minimal cardinality of a set of edges of G, if exists, whose deletion disconnects G and each remaining component contains more than g vertices. In this paper, we show that the n-dimensional balanced hypercube B H n, which is a variant of hypercube Q n, is still strongly Menger (edge) connected even when there are ( 2 n − 4 ) faulty vertices (resp. ( 2 n − 2 ) faulty edges) for n ≥ 2. Moreover, under the restricted condition that each vertex has at least two fault-free edges, we show that B H n is still strongly Menger edge connected even when there are ( 6 n − 8 ) faulty edges for n ≥ 2. These results are all optimal with respect to the maximum number of tolerated (edge) faults. Furthermore, we showed that the 3-extra edge-connectivity of B H n is 8 n − 8 for n ≥ 2.

Details DOI

AAAI Conference 2018 Conference Paper

Multi-Rate Gated Recurrent Convolutional Networks for Video-Based Pedestrian Re-Identification

  • Zhihui Li
  • Lina Yao
  • Feiping Nie
  • Dingwen Zhang
  • Min Xu

Matching pedestrians across multiple camera views has attracted lots of recent research attention due to its apparent importance in surveillance and security applications. While most existing works address this problem in a still-image setting, we consider the more informative and challenging video-based person re-identification problem, where a video of a pedestrian as seen in one camera needs to be matched to a gallery of videos captured by other non-overlapping cameras. We employ a convolutional network to extract the appearance and motion features from raw video sequences, and then feed them into a multi-rate recurrent network to exploit the temporal correlations, and more importantly, to take into account the fact that pedestrians, sometimes even the same pedestrian, move in different speeds across different camera views. The combined network is trained in an end-to-end fashion, and we further propose an initialization strategy via context reconstruction to largely improve the performance. We conduct extensive experiments on the iLIDS-VID and PRID-2011 datasets, and our experimental results confirm the effectiveness and the generalization ability of our model.

PDF Details

TCS Journal 2017 Journal Article

Conditional diagnosability of a class of matching composition networks under the comparison model

  • Min Xu
  • Krishnaiyan Thulasiraman
  • Qiang Zhu

Fault diagnosis of interconnection networks is an important consideration in the design and maintenance of multiprocessor systems. Herein, we study fault diagnosis, which is the identification of faulty processors in high speed parallel processing systems. Conditional diagnosability, proposed by Lai et al. [22], assumes that no fault set can contain all the neighbors of any processor in a system; this is a well-accepted and general measure of the diagnosis ability of an interconnection network of multiprocessor systems. The diagnosability and conditional diagnosability of many interconnection networks have been studied using various diagnosis models. In this paper we study the conditional diagnosability of matching composition networks under the comparison model (MM* model). In [31] Yang determined a set of sufficient conditions for a network G to be conditionally ( 3 n − 3 − C ( G ) ) -diagnosable. Our main contribution in this paper is to extend Yang's result by determining a larger class of networks that are conditionally ( 3 n − 3 − C ( G ) ) -diagnosable. Yang's result [31] and earlier results for the hypercube, the crossed cube, the twisted cube and the Möbius cube [18, 32, 33] all become corollaries of our main result. Thus this paper extends the state of the art in the area of conditional diagnosability of multiprocessor systems.

Details DOI

IJCAI Conference 2017 Conference Paper

Dependency Exploitation: A Unified CNN-RNN Approach for Visual Emotion Recognition

  • Xinge Zhu
  • Liang Li
  • Weigang Zhang
  • Tianrong Rao
  • Min Xu
  • Qingming Huang
  • Dong Xu

Visual emotion recognition aims to associate images with appropriate emotions. There are different visual stimuli that can affect human emotion from low-level to high-level, such as color, texture, part, object, etc. However, most existing methods treat different levels of features as independent entity without having effective method for feature fusion. In this paper, we propose a unified CNN-RNN model to predict the emotion based on the fused features from different levels by exploiting the dependency among them. Our proposed architecture leverages convolutional neural network (CNN) with multiple layers to extract different levels of features with in a multi-task learning framework, in which two related loss functions are introduced to learn the feature representation. Considering the dependencies within the low-level and high-level features, a new bidirectional recurrent neural network (RNN) is proposed to integrate the learned features from different layers in the CNN model. Extensive experiments on both Internet images and art photo datasets demonstrate that our method outperforms the state-of-the-art methods with at least 7% performance improvement.

PDF Details

TCS Journal 2017 Journal Article

On g-good-neighbor conditional diagnosability of (n,k)-star networks

  • Yulong Wei
  • Min Xu

The g-good-neighbor conditional diagnosability is a new measure for fault diagnosis of systems. Xu et al. (2017) [27] determined the g-good-neighbor conditional diagnosability of ( n, k ) -star networks S n, k (i. e. , t g ( S n, k ) ) with 1 ≤ k ≤ n − 1 for 1 ≤ g ≤ n − k under the PMC model and the MM⁎ model. In this paper, we determine t g ( S n, k ) for all the remaining cases with 1 ≤ k ≤ n − 1 for 1 ≤ g ≤ n − 1 under the two models, from which we can obtain the g-good-neighbor conditional diagnosability of the star graph obtained by Li et al. (2017) [16] for 1 ≤ g ≤ n − 2.

Details DOI

NeurIPS Conference 2013 Conference Paper

Estimation Bias in Multi-Armed Bandit Algorithms for Search Advertising

  • Min Xu
  • Tao Qin
  • Tie-Yan Liu

In search advertising, the search engine needs to select the most profitable advertisements to display, which can be formulated as an instance of online learning with partial feedback, also known as the stochastic multi-armed bandit (MAB) problem. In this paper, we show that the naive application of MAB algorithms to search advertising for advertisement selection will produce sample selection bias that harms the search engine by decreasing expected revenue and “estimation of the largest mean” (ELM) bias that harms the advertisers by increasing game-theoretic player-regret. We then propose simple bias-correction methods with benefits to both the search engine and the advertisers.

PDF Details

JMLR Journal 2011 Journal Article

Forest Density Estimation

  • Han Liu
  • Min Xu
  • Haijie Gu
  • Anupam Gupta
  • John Lafferty
  • Larry Wasserman

We study graph estimation and density estimation in high dimensions, using a family of density estimators based on forest structured undirected graphical models. For density estimation, we do not assume the true distribution corresponds to a forest; rather, we form kernel density estimates of the bivariate and univariate marginals, and apply Kruskal's algorithm to estimate the optimal forest on held out data. We prove an oracle inequality on the excess risk of the resulting estimator relative to the risk of the best forest. For graph estimation, we consider the problem of estimating forests with restricted tree sizes. We prove that finding a maximum weight spanning forest with restricted tree size is NP-hard, and develop an approximation algorithm for this problem. Viewing the tree size as a complexity parameter, we then select a forest using data splitting, and prove bounds on excess risk and structure selection consistency of the procedure. Experiments with simulated data and microarray data indicate that the methods are a practical alternative to Gaussian graphical models. [abs] [ pdf ][ bib ] &copy JMLR 2011. ( edit, beta )

PDF Details

JMLR Journal 2011 Journal Article

High-dimensional Covariance Estimation Based On Gaussian Graphical Models

  • Shuheng Zhou
  • Philipp Rütimann
  • Min Xu
  • Peter Bühlmann

Undirected graphs are often used to describe high dimensional distributions. Under sparsity conditions, the graph can be estimated using l 1 -penalization methods. We propose and study the following method. We combine a multiple regression approach with ideas of thresholding and refitting: first we infer a sparse undirected graphical model structure via thresholding of each among many l 1 -norm penalized regression functions; we then estimate the covariance matrix and its inverse using the maximum likelihood estimator. We show that under suitable conditions, this approach yields consistent estimation in terms of graphical structure and fast convergence rates with respect to the operator and Frobenius norm for the covariance matrix and its inverse. We also derive an explicit bound for the Kullback Leibler divergence. [abs] [ pdf ][ bib ] &copy JMLR 2011. ( edit, beta )

PDF Details

NeurIPS Conference 2011 Conference Paper

Noise Thresholds for Spectral Clustering

  • Sivaraman Balakrishnan
  • Min Xu
  • Akshay Krishnamurthy
  • Aarti Singh

Although spectral clustering has enjoyed considerable empirical success in machine learning, its theoretical properties are not yet fully developed. We analyze the performance of a spectral algorithm for hierarchical clustering and show that on a class of hierarchically structured similarity matrices, this algorithm can tolerate noise that grows with the number of data points while still perfectly recovering the hierarchical clusters with high probability. We additionally improve upon previous results for k-way spectral clustering to derive conditions under which spectral clustering makes no mistakes. Further, using minimax analysis, we derive tight upper and lower bounds for the clustering problem and compare the performance of spectral clustering to these information theoretic limits. We also present experiments on simulated and real world data illustrating our results.

PDF Details

AIIM Journal 2007 Journal Article

Advanced fuzzy cellular neural network: Application to CT liver images

  • Shitong Wang
  • Duan Fu
  • Min Xu
  • Dewen Hu

Objective To achieve better boundary integrities and recall accuracies for segmented liver images, use of the advanced fuzzy cellular neural network (AFCNN), as a variant of the fuzzy cellular neural network (FCNN), is proposed to effectively segment CT liver images. Materials and methods In order to better utilize relevant contour and gray information from liver images, we have improved the FCNN [Wang S, Wang M. A new algorithm NDA based on fuzzy cellular neural networks for white blood cell detection. IEEE Trans Inform Technol Biomed, in press], which proved to be very effective for the segmentation of microscopic white blood cell images, to create the novel neural network, AFCNN. Its convergent property and global stability are proved. Based on the FCNN-based NDA algorithm [Wang S, Wang M. A new algorithm NDA based on fuzzy cellular neural networks for white blood cell detection. IEEE Trans Inform Technol Biomed, in press], we developed the AFCNN-based NDA algorithm, which we used to segment 5 CT liver images. For comparison, we also segmented the same 5 CT liver images using the FCNN-based NDA algorithm. Results and conclusion: AFCNN has distinct advantages over FCNN in both boundary integrity and recall accuracy. In particular, the performance index Binary_rate is generally much higher for AFCNN than for FCNN when applied to CT liver images.

Details DOI

KER Journal 2003 Journal Article

OntoVote: a scalable distributed vote-collecting mechanism for ontology drift on a P2P platform

  • YANFENG GE
  • Yong Yu
  • Xing Zhu
  • Shen Huang
  • Min Xu

Ontologies provide potential support for knowledge and content management on a P2P platform. Although we can design ontologies beforehand for an application, it is argued that in P2P environments static or predefined ontologies cannot satisfy the ever-changing requirements of all users. So we propose every user should make proposals for what kind of ontology is the most apt to his need. Collecting all these proposals (or votes) helps the drift of ontologies. This paper introduces OntoVote, a scalable distributed vote-collecting mechanism based on application-level broadcast trees, and describes how OntoVote can be applied to ontology drift on a P2P platform by discussing several problems involved in the voting process.

Details DOI
v2026.09.13