arXiv subjects

Chunlei Li

Publications and source records attributed to Chunlei Li.

At least 37 records · Page 2Linked to original sources

arXiv2025

Permutation polynomials over finite fields from low-degree rational functions

This paper considers permutation polynomials over the finite field $F_{q^2}$ in even characteristic by utilizing low-degree permutation rational functions over $F_q$. As a result, we obtain two classes of permutation binomials and six classes of permutation pentanomials over $F_{q^2}$. Additionally, we show that the obtained binomials and pentanomials are quasi-multiplicative inequivalent to the known ones in the literature.

cs.CR↗

arXiv2025

Multimodal Large Language Models for Medical Report Generation via Customized Prompt Tuning

Medical report generation from imaging data remains a challenging task in clinical practice. While large language models (LLMs) show great promise in addressing this challenge, their effective integration with medical imaging data still deserves in-depth exploration. In this paper, we present MRG-LLM, a novel multimodal large language model (MLLM) that combines a frozen LLM with a learnable visual encoder and introduces a dynamic prompt customization mechanism. Our key innovation lies in generating instance-specific prompts tailored to individual medical images through conditional affine transformations derived from visual features. We propose two implementations: prompt-wise and promptbook-wise customization, enabling precise and targeted report generation. Extensive experiments on IU X-ray and MIMIC-CXR datasets demonstrate that MRG-LLM achieves state-of-the-art performance in medical report generation. Our code will be made publicly available.

cs.CV↗

arXiv2025

Taming Stable Diffusion for Computed Tomography Blind Super-Resolution

High-resolution computed tomography (CT) imaging is essential for medical diagnosis but requires increased radiation exposure, creating a critical trade-off between image quality and patient safety. While deep learning methods have shown promise in CT super-resolution, they face challenges with complex degradations and limited medical training data. Meanwhile, large-scale pre-trained diffusion models, particularly Stable Diffusion, have demonstrated remarkable capabilities in synthesizing fine details across various vision tasks. Motivated by this, we propose a novel framework that adapts Stable Diffusion for CT blind super-resolution. We employ a practical degradation model to synthesize realistic low-quality images and leverage a pre-trained vision-language model to generate corresponding descriptions. Subsequently, we perform super-resolution using Stable Diffusion with a specialized controlling strategy, conditioned on both low-resolution inputs and the generated text descriptions. Extensive experiments show that our method outperforms existing approaches, demonstrating its potential for achieving high-quality CT imaging at reduced radiation doses. Our code will be made publicly available.

eess.IV↗

arXiv2025

Constructions of Optimal Frequency-Hopping Sequences with Controlled Minimum Gaps

Frequency-hopping sequences (FHSs) with low Hamming correlation and wide gaps significantly contribute to the anti-interference performance in FH communication systems. This paper investigates FHSs with optimal Hamming correlation and controlled minimum gaps. We start with the discussion of the upper bounds on the minimum gaps of uniform FHSs and then propose a general construction of optimal uniform wide-gap FHSs with length 2l and 3l, which includes the work by Li et al. in IEEE Trans. Inf. Theory, vol. 68, no. 1, 2022 as a special case. Furthermore, we present a recursive construction of FHSs with length 2l, which concatenate shorter sequences of known minimum gaps. It is shown that the resulting FHSs have the same Hamming correlation as the concatenation-ordering sequences. As applications, several known optimal FHSs are used to produce optimal FHSs with controlled minimum gaps.

cs.IT↗

arXiv2025

MGPBD: A Multigrid Accelerated Global XPBD Solver

We introduce a novel Unsmoothed Aggregation (UA) Algebraic Multigrid (AMG) method combined with Preconditioned Conjugate Gradient (PCG) to overcome the limitations of Extended Position-Based Dynamics (XPBD) in high-resolution and high-stiffness simulations. While XPBD excels in simulating deformable objects due to its speed and simplicity, its nonlinear Gauss-Seidel (GS) solver often struggles with low-frequency errors, leading to instability and stalling issues, especially in high-resolution, high-stiffness simulations. Our multigrid approach addresses these issues efficiently by leveraging AMG. To reduce the computational overhead of traditional AMG, where prolongator construction can consume up to two-thirds of the runtime, we propose a lazy setup strategy that reuses prolongators across iterations based on matrix structure and physical significance. Furthermore, we introduce a simplified method for constructing near-kernel components by applying a few sweeps of iterative methods to the homogeneous equation, achieving convergence rates comparable to adaptive smoothed aggregation (adaptive-SA) at a lower computational cost. Experimental results demonstrate that our method significantly improves convergence rates and numerical stability, enabling efficient and stable high-resolution simulations of deformable objects.

cs.GR↗

arXiv2025

UniCrossAdapter: Multimodal Adaptation of CLIP for Radiology Report Generation

Automated radiology report generation aims to expedite the tedious and error-prone reporting process for radiologists. While recent works have made progress, learning to align medical images and textual findings remains challenging due to the relative scarcity of labeled medical data. For example, datasets for this task are much smaller than those used for image captioning in computer vision. In this work, we propose to transfer representations from CLIP, a large-scale pre-trained vision-language model, to better capture cross-modal semantics between images and texts. However, directly applying CLIP is suboptimal due to the domain gap between natural images and radiology. To enable efficient adaptation, we introduce UniCrossAdapter, lightweight adapter modules that are incorporated into CLIP and fine-tuned on the target task while keeping base parameters fixed. The adapters are distributed across modalities and their interaction to enhance vision-language alignment. Experiments on two public datasets demonstrate the effectiveness of our approach, advancing state-of-the-art in radiology report generation. The proposed transfer learning framework provides a means of harnessing semantic knowledge from large-scale pre-trained models to tackle data-scarce medical vision-language tasks. Code is available at https://github.com/chauncey-tow/MRG-CLIP.

cs.CV↗

arXiv2025

Reducing Annotation Burden: Exploiting Image Knowledge for Few-Shot Medical Video Object Segmentation via Spatiotemporal Consistency Relearning

Few-shot video object segmentation aims to reduce annotation costs; however, existing methods still require abundant dense frame annotations for training, which are scarce in the medical domain. We investigate an extremely low-data regime that utilizes annotations from only a few video frames and leverages existing labeled images to minimize costly video annotations. Specifically, we propose a two-phase framework. First, we learn a few-shot segmentation model using labeled images. Subsequently, to improve performance without full supervision, we introduce a spatiotemporal consistency relearning approach on medical videos that enforces consistency between consecutive frames. Constraints are also enforced between the image model and relearning model at both feature and prediction levels. Experiments demonstrate the superiority of our approach over state-of-the-art few-shot segmentation methods. Our model bridges the gap between abundant annotated medical images and scarce, sparsely labeled medical videos to achieve strong video segmentation performance in this low data regime. Code is available at https://github.com/MedAITech/RAB.

cs.CV↗

arXiv2025

One-Shot Medical Video Object Segmentation via Temporal Contrastive Memory Networks

Video object segmentation is crucial for the efficient analysis of complex medical video data, yet it faces significant challenges in data availability and annotation. We introduce the task of one-shot medical video object segmentation, which requires separating foreground and background pixels throughout a video given only the mask annotation of the first frame. To address this problem, we propose a temporal contrastive memory network comprising image and mask encoders to learn feature representations, a temporal contrastive memory bank that aligns embeddings from adjacent frames while pushing apart distant ones to explicitly model inter-frame relationships and stores these features, and a decoder that fuses encoded image features and memory readouts for segmentation. We also collect a diverse, multi-source medical video dataset spanning various modalities and anatomies to benchmark this task. Extensive experiments demonstrate state-of-the-art performance in segmenting both seen and unseen structures from a single exemplar, showing ability to generalize from scarce labels. This highlights the potential to alleviate annotation burdens for medical video analysis. Code is available at https://github.com/MedAITech/TCMN.

cs.CV↗

arXiv2025

Scale-Aware Contrastive Reverse Distillation for Unsupervised Medical Anomaly Detection

Unsupervised anomaly detection using deep learning has garnered significant research attention due to its broad applicability, particularly in medical imaging where labeled anomalous data are scarce. While earlier approaches leverage generative models like autoencoders and generative adversarial networks (GANs), they often fall short due to overgeneralization. Recent methods explore various strategies, including memory banks, normalizing flows, self-supervised learning, and knowledge distillation, to enhance discrimination. Among these, knowledge distillation, particularly reverse distillation, has shown promise. Following this paradigm, we propose a novel scale-aware contrastive reverse distillation model that addresses two key limitations of existing reverse distillation methods: insufficient feature discriminability and inability to handle anomaly scale variations. Specifically, we introduce a contrastive student-teacher learning approach to derive more discriminative representations by generating and exploring out-of-normal distributions. Further, we design a scale adaptation mechanism to softly weight contrastive distillation losses at different scales to account for the scale variation issue. Extensive experiments on benchmark datasets demonstrate state-of-the-art performance, validating the efficacy of the proposed method. Code is available at https://github.com/MedAITech/SCRD4AD.

cs.CV↗

arXiv2025

Rethinking Cell Counting Methods: Decoupling Counting and Localization

Cell counting in microscopy images is vital in medicine and biology but extremely tedious and time-consuming to perform manually. While automated methods have advanced in recent years, state-of-the-art approaches tend to increasingly complex model designs. In this paper, we propose a conceptually simple yet effective decoupled learning scheme for automated cell counting, consisting of separate counter and localizer networks. In contrast to jointly learning counting and density map estimation, we show that decoupling these objectives surprisingly improves results. The counter operates on intermediate feature maps rather than pixel space to leverage global context and produce count estimates, while also generating coarse density maps. The localizer then reconstructs high-resolution density maps that precisely localize individual cells, conditional on the original images and coarse density maps from the counter. Besides, to boost counting accuracy, we further introduce a global message passing module to integrate cross-region patterns. Extensive experiments on four datasets demonstrate that our approach, despite its simplicity, challenges common practice and achieves state-of-the-art performance by significant margins. Our key insight is that decoupled learning alleviates the need to learn counting on high-resolution density maps directly, allowing the model to focus on global features critical for accurate estimates. Code is available at https://github.com/MedAITech/DCL.

cs.CV↗

arXiv2025

Bounded-degree Low Rank Parity Check Codes

Low-rank parity-check (LRPC) codes are the rank-metric analogue of low-density parity-check codes and they found important applications in code-based cryptography. In this paper we investigate a sub-family of LRPC codes, which have a parity-check matrix defined over a subspace $\calV_{α,d}=\Span{\Fq}{1,α, \ldots, α^{d-1}}\subsetneq \Fqm$, where $\Fqm$ is the finite field of $q^m$ elements and $d$ is a positive integer significantly smaller than $m $; and they are termed bounded-degree LRPC (BD-LRPC) codes. These codes are the same as the standard LRPC codes of density $2$ when the degree $d=2$, while for degree $d>2$ they constitute a proper subset of LRPC codes of density $d$. Exploiting the structure of $\calV_{α,d}$, the BD-LRPC codes of degree $d$ can uniquely correct errors of rank weight $r$ when $n-k \geq r + u$ for certain $u \geq 1$, in contrast to the condition $n-k\geq dr$ required for the standard LRPC codes. This underscores the superior decoding capability of the BD-LRPC codes. Moreover, as the code length $n\rightarrow \infty$, when $n/m\rightarrow 0$, the BD-LRPC codes with a code rate of $R=k/n$ can be uniquely decodable with radius $ρ=r/n$ approaching the Singleton bound $1-R$ by letting $ε=u/n\rightarrow 0$; and when $n/m$ is a constant, the BD-LRPC codes can have unique decoding radius $ρ= 1-R-ε$ for a small $ε$, allowing for $ρ>(1-R)/2$ with properly chosen parameters. This superior decoding capability is theoretically proved for the case $d=2$ and confirmed by experimental results for $d>2$.

cs.IT↗

arXiv2024

Multi-feature Reconstruction Network using Crossed-mask Restoration for Unsupervised Industrial Anomaly Detection

Unsupervised anomaly detection using only normal samples is of great significance for quality inspection in industrial manufacturing. Although existing reconstruction-based methods have achieved promising results, they still face two problems: poor distinguishable information in image reconstruction and well abnormal regeneration caused by model under-regularization. To overcome the above issues, we convert the image reconstruction into a combination of parallel feature restorations and propose a multi-feature reconstruction network, MFRNet, using crossed-mask restoration in this paper. Specifically, a multi-scale feature aggregator is first developed to generate more discriminative hierarchical representations of the input images from a pre-trained model. Subsequently, a crossed-mask generator is adopted to randomly cover the extracted feature map, followed by a restoration network based on the transformer structure for high-quality repair of the missing regions. Finally, a hybrid loss is equipped to guide model training and anomaly estimation, which gives consideration to both the pixel and structural similarity. Extensive experiments show that our method is highly competitive with or significantly outperforms other state-of-the-arts on four public available datasets and one self-made dataset.

cs.CV↗

arXiv2024

Enhanced Anomaly Detection in Industrial Control Systems aided by Machine Learning

Traditional intrusion detection systems (IDSs) often rely on either network traffic or process data, but this single-source approach may miss complex attack patterns that span multiple layers within industrial control systems (ICSs) or persistent threats that target different layers of operational technology systems. This study investigates whether combining both network and process data can improve attack detection in ICSs environments. Leveraging the SWaT dataset, we evaluate various machine learning models on individual and combined data sources. Our findings suggest that integrating network traffic with operational process data can enhance detection capabilities, evidenced by improved recall rates for cyber attack classification. Serving as a proof-of-concept within a limited testing environment, this research explores the feasibility of advancing intrusion detection through a multi-source data approach in ICSs. Although the results are promising, they are preliminary and highlight the need for further studies across diverse datasets and refined methodologies.

cs.CR↗

arXiv2024

A Further Investigation on Complete Complementary Codes from $q$-ary Functions

This research focuses on constructing $q$-ary functions for complete complementary codes (CCCs) with flexible parameters. Most existing work has primarily identified sufficient conditions for $q$-ary functions related to $q$-ary CCCs. To the best of the authors' knowledge, this study is the first to establish both the necessary and sufficient conditions for $q$-ary functions, encompassing most existing CCCs constructions as special cases. For $q$-ary CCCs with a length of $q^m$ and a set size of $q^{n+1}$, we begin by analyzing the necessary and sufficient conditions for $q$-ary functions defined over the domain $\mathbb{Z}_q^m$. Additionally, we construct CCCs with lengths given by $L = \prod_{i=1}^k p_i^{m_i}$, set sizes given by $K = \prod_{i=1}^k p_i^{n_i+1}$, and an alphabet size of $ν= \prod_{i=1}^k p_i$, where $p_1 < p_2 < \cdots < p_k$. To achieve these specific parameters, we examine the necessary and sufficient conditions for $ν$-ary functions over the domain $\mathbf{Z}_{p_1}^{m_1} \times \cdots \times \mathbf{Z}_{p_k}^{m_k}$, which is a subset of $\mathbb{Z}_ν^m$ and contains $\prod_{i=1}^k p_i^{m_i}$ vectors. In this context, $\mathbf{Z}_{p_i}^{m_i} = \{0, 1, \ldots, p_i - 1\}^{m_i}$, and $m$ is the sum of $m_1, m_2, \ldots, m_k$. The $q$-ary and $ν$-ary functions allow us to cover all possible length sequences. However, we find that the proposed $ν$-ary functions are more suitable for generating CCCs with a length of $L = \prod_{i=1}^k p_i^{m_i}$, particularly when $m_i$ is coprime to $m_j$ for some $1 \leq i \neq j \leq k$. While the proposed $q$-ary functions can also produce CCCs of the same length $L$, the set size and alphabet size become as large as $L$, since in this case, the only choice for $q$ is $L$. In contrast, the proposed $ν$-ary functions yield CCCs with a more flexible set size $K\leq L$ and an alphabet size of $ν<L$.

math.CO↗

arXiv2024

Further Investigation on Differential Properties of the Generalized Ness-Helleseth Function

Let $n$ be an odd positive integer, $p$ be a prime with $p\equiv3\pmod4$, $d_{1} = {{p^{n}-1}\over {2}} -1 $ and $d_{2} =p^{n}-2$. The function defined by $f_u(x)=ux^{d_{1}}+x^{d_{2}}$ is called the generalized Ness-Helleseth function over $\mathbb{F}_{p^n}$, where $u\in\mathbb{F}_{p^n}$. It was initially studied by Ness and Helleseth in the ternary case. In this paper, for $p^n \equiv 3 \pmod 4$ and $p^n \ge7$, we provide the necessary and sufficient condition for $f_u(x)$ to be an APN function. In addition, for each $u$ satisfying $χ(u+1) = χ(u-1)$, the differential spectrum of $f_u(x)$ is investigated, and it is expressed in terms of some quadratic character sums of cubic polynomials, where $χ(\cdot)$ denotes the quadratic character of $\mathbb{F}_{p^n}$.

cs.CR↗

arXiv2024

An updated review on cross-correlation of m-sequences

Maximum-length sequences (m-sequences for short) over finite fields are generated by linear feedback shift registers with primitive characteristic polynomials. These sequences have nice mathematical structures and good randomness properties that are favorable in practical applications. During the past five decades, the crosscorrelation between m-sequences of the same period has been intensively studied, and a particular research focus has been on investigating the cross-correlation spectra with few possibles values. In this chapter we summarize all known results on this topic in the literature and promote several open problems for future research.

cs.CR↗

arXiv2024

Further Investigations on Nonlinear Complexity of Periodic Binary Sequences

Nonlinear complexity is an important measure for assessing the randomness of sequences. In this paper we investigate how circular shifts affect the nonlinear complexities of finite-length binary sequences and then reveal a more explicit relation between nonlinear complexities of finite-length binary sequences and their corresponding periodic sequences. Based on the relation, we propose two algorithms that can generate all periodic binary sequences with any prescribed nonlinear complexity.

cs.IT↗

arXiv2024

Construction of Complete Complementary Codes over Small Alphabet

Complete complementary codes (CCCs) play a vital role not only in wireless communication, particularly in multicarrier systems where achieving an interference-free environment is of paramount importance, but also in the construction of other codes that necessitate appropriate functions to meet the diverse demands within today's landscape of wireless communication evaluation. This research is focused on the area of constructing $q$-ary functions for both of {traditional and spectrally null constraint (SNC) CCCs}\footnote{When no codes in CCCs having zero components, we call it as traditonal CCCs, else, we call it as SNC-CCCs in this pape.} of flexible length, set size and alphabet. We construct traditional CCCs with lengths, defined as $L = \prod_{i=1}^k p_i^{m_i}$, set sizes, defined as $K = \prod_{i=1}^k p_i^{n_i+1}$, and an alphabet size of $q=\prod_{i=1}^k p_i$, such that $p_1<p_2<\cdots<p_k $. The parameters $m_1, m_2, \ldots, m_k$ (each greater than or equal to $2$) are positive integers, while $n_1, n_2, \ldots, n_k$ are non-negative integers satisfying $n_i \leq m_i-1$, and the variable $k$ represents a positive integer. To achieve these specific parameters, we define $q$-ary functions over a domain $\mathbf{Z}_{p_1}^{m_1}\times \cdots \times \mathbf{Z}_{p_k}^{m_k}$ that is considered a proper subset of $\mathbb{Z}_{q}^m$ and encompasses $\prod_{i=1}^k p_i^{m_i}$ vectors, where $\mathbf{Z}_{p_i}^{m_i}=\{0,1,\hdots,p_i-1\}^{m_i}$, and the value of $m$ is derived from the sum of $m_1, m_2, \ldots, m_k$. This organization of the domain allows us to encompass all conceivable integer-valued length sequences over the alphabet $\mathbb{Z}_q$. It has been demonstrated that by constraining a $q$-ary function that generates traditional CCCs, we can derive SNC-CCCs with identical length and alphabet, yet a smaller or equal set size compared to the traditional CCCs.

cs.IT↗