Searcharxiv⌕ Search

arXiv subjects

Wenhua Qian

Publications and source records attributed to Wenhua Qian.

13 recordsLinked to original sources
arXiv2026

Multi-Modal Image Fusion via Intervention-Stable Feature Learning

Multi-modal image fusion integrates complementary information from different modalities into a unified representation. Current methods predominantly optimize statistical correlations between modalities, often capturing dataset-induced spurious associations that degrade under distribution shifts. In this paper, we propose an intervention-based framework inspired by causal principles to identify robust cross-modal dependencies. Drawing insights from Pearl's causal hierarchy, we design three principled intervention strategies to probe different aspects of modal relationships: i) complementary masking with spatially disjoint perturbations tests whether modalities can genuinely compensate for each other's missing information, ii) random masking of identical regions identifies feature subsets that remain informative under partial observability, and iii) modality dropout evaluates the irreplaceable contribution of each modality. Based on these interventions, we introduce a Causal Feature Integrator (CFI) that learns to identify and prioritize intervention-stable features maintaining importance across different perturbation patterns through adaptive invariance gating, thereby capturing robust modal dependencies rather than spurious correlations. Extensive experiments demonstrate that our method achieves SOTA performance on both public benchmarks and downstream high-level vision tasks.

cs.CV↗
arXiv2025

Residual Prior-driven Frequency-aware Network for Image Fusion

Image fusion aims to integrate complementary information across modalities to generate high-quality fused images, thereby enhancing the performance of high-level vision tasks. While global spatial modeling mechanisms show promising results, constructing long-range feature dependencies in the spatial domain incurs substantial computational costs. Additionally, the absence of ground-truth exacerbates the difficulty of capturing complementary features effectively. To tackle these challenges, we propose a Residual Prior-driven Frequency-aware Network, termed as RPFNet. Specifically, RPFNet employs a dual-branch feature extraction framework: the Residual Prior Module (RPM) extracts modality-specific difference information from residual maps, thereby providing complementary priors for fusion; the Frequency Domain Fusion Module (FDFM) achieves efficient global feature modeling and integration through frequency-domain convolution. Additionally, the Cross Promotion Module (CPM) enhances the synergistic perception of local details and global structures through bidirectional feature interaction. During training, we incorporate an auxiliary decoder and saliency structure loss to strengthen the model's sensitivity to modality-specific differences. Furthermore, a combination of adaptive weight-based frequency contrastive loss and SSIM loss effectively constrains the solution space, facilitating the joint capture of local details and global features while ensuring the retention of complementary information. Extensive experiments validate the fusion performance of RPFNet, which effectively integrates discriminative features, enhances texture details and salient objects, and can effectively facilitate the deployment of the high-level vision task.

cs.CV↗
arXiv2025

Coarse-to-Fine Structure-Aware Artistic Style Transfer

Artistic style transfer aims to use a style image and a content image to synthesize a target image that retains the same artistic expression as the style image while preserving the basic content of the content image. Many recently proposed style transfer methods have a common problem; that is, they simply transfer the texture and color of the style image to the global structure of the content image. As a result, the content image has a local structure that is not similar to the local structure of the style image. In this paper, we present an effective method that can be used to transfer style patterns while fusing the local style structure into the local content structure. In our method, dif-ferent levels of coarse stylized features are first reconstructed at low resolution using a Coarse Network, in which style color distribution is roughly transferred, and the content structure is combined with the style structure. Then, the reconstructed features and the content features are adopted to synthesize high-quality structure-aware stylized images with high resolution using a Fine Network with three structural selective fusion (SSF) modules. The effectiveness of our method is demonstrated through the generation of appealing high-quality stylization results and a com-parison with some state-of-the-art style transfer methods.

cs.CV↗
arXiv2023

LP-BFGS attack: An adversarial attack based on the Hessian with limited pixels

Deep neural networks are vulnerable to adversarial attacks. Most $L_{0}$-norm based white-box attacks craft perturbations by the gradient of models to the input. Since the computation cost and memory limitation of calculating the Hessian matrix, the application of Hessian or approximate Hessian in white-box attacks is gradually shelved. In this work, we note that the sparsity requirement on perturbations naturally lends itself to the usage of Hessian information. We study the attack performance and computation cost of the attack method based on the Hessian with a limited number of perturbation pixels. Specifically, we propose the Limited Pixel BFGS (LP-BFGS) attack method by incorporating the perturbation pixel selection strategy and the BFGS algorithm. Pixels with top-k attribution scores calculated by the Integrated Gradient method are regarded as optimization variables of the LP-BFGS attack. Experimental results across different networks and datasets demonstrate that our approach has comparable attack ability with reasonable computation in different numbers of perturbation pixels compared with existing solutions.

cs.CR↗
arXiv2022

Joint spectrum shrinking maps on projections

Let $\mathcal H$ be a finite dimensional complex Hilbert space with dimension $n \ge 3$ and $\mathcal P(\mathcal H)$ the set of projections on $\mathcal H$. Let $φ: \mathcal P(\mathcal H) \to \mathcal P(\mathcal H)$ be a surjective map. We show that $φ$ shrinks the joint spectrum of any two projections if and only if it is joint spectrum preserving for any two projections and thus is induced by a ring automorphism on $\mathbb C$ in a particular way. In addition, for an arbitrary $k \ge 3$, $φ$ shrinks the joint spectrum of any $k$ projections if and only if it is induced by a unitary or an anti-unitary. Assume that $ϕ$ is a surjective map on the Grassmann space of rank one projections. We show that $ϕ$ is joint spectrum preserving for any $n$ rank one projections if and only if it can be extended to a surjective map on $\mathcal P(\mathcal{H})$ which is spectrum preserving for any two projections. Moreover, for any $k >n$, $ϕ$ is joint spectrum shrinking for any $k$ rank one projections if and only if it is induced by a unitary or an anti-unitary.

math.FA↗
arXiv2022

Exploring Adversarial Examples and Adversarial Robustness of Convolutional Neural Networks by Mutual Information

A counter-intuitive property of convolutional neural networks (CNNs) is their inherent susceptibility to adversarial examples, which severely hinders the application of CNNs in security-critical fields. Adversarial examples are similar to original examples but contain malicious perturbations. Adversarial training is a simple and effective defense method to improve the robustness of CNNs to adversarial examples. The mechanisms behind adversarial examples and adversarial training are worth exploring. Therefore, this work investigates similarities and differences between normally trained CNNs (NT-CNNs) and adversarially trained CNNs (AT-CNNs) in information extraction from the mutual information perspective. We show that 1) whether NT-CNNs or AT-CNNs, for original and adversarial examples, the trends towards mutual information are almost similar throughout training; 2) compared with normal training, adversarial training is more difficult and the amount of information that AT-CNNs extract from the input is less; 3) the CNNs trained with different methods have different preferences for certain types of information; NT-CNNs tend to extract texture-based information from the input, while AT-CNNs prefer to shape-based information. The reason why adversarial examples mislead CNNs may be that they contain more texture-based information about other classes. Furthermore, we also analyze the mutual information estimators used in this work and find that they outline the geometric properties of the middle layer's output.

cs.LG↗
arXiv2021

Surjective $L^p$-isometries of Grassmann spaces

Based on the characterization of surjective $L^p$-isometries of unitary groups in finite factors, we describe all surjective $L^p$-isometries between Grassmann spaces of projections with the same trace value in semifinite factors.

math.OA↗
arXiv2018

Wigner-Type Theorem on transition probability preserving maps in semifinite factors

The Wigner's theorem, which is one of the cornerstones of the mathematical formulation of quantum mechanics, asserts that every symmetry of quantum system is unitary or anti-unitary. This classical result was first given by Wigner in 1931. Thereafter it has been proved and generalized in various ways by many authors. Recently, G. P. Gehér extended Wigner's and Molnár's theorems and characterized the transformations on the Grassmann space of all rank-$n$ projections which preserve the transition probability. The aim of this paper is to provide a new approach to describe the general form of the transition probability preserving (not necessarily bijective) maps between Grassmann spaces. As a byproduct, we are able to generalize the results of Molnár and G. P. Gehér.

math.OA↗
arXiv2018

Derivations and 2-local derivations on matrix algebras over commutative algebras

We characterize derivations and 2-local derivations from $M_{n}(\mathcal{A})$ into $M_{n}(\mathcal{M})$, $n \ge 2$, where $\mathcal{A}$ is a unital algebra over $\mathbb{C}$ and $\mathcal{M}$ is a unital $\mathcal{A}$-bimodule. We show that every derivation $D: M_{n}(\mathcal{A}) \to M_{n}(\mathcal{M})$, $n \ge 2,$ is the sum of an inner derivation and a derivation induced by a derivation from $\mathcal{A}$ to $\mathcal{M}$. We say that $\mathcal{A}$ commutes with $\mathcal{M}$ if $am=ma$ for every $a\in\mathcal{A}$ and $m\in\mathcal{M}$. If $\mathcal{A}$ commutes with $\mathcal{M}$ we prove that every inner 2-local derivation $D: M_{n}(\mathcal{A}) \to M_{n}(\mathcal{M})$, $n \ge 2$, is an inner derivation. In addition, if $\mathcal{A}$ is commutative and commutes with $\mathcal{M}$, then every 2-local derivation $D: M_{n}(\mathcal{A}) \to M_{n}(\mathcal{M})$, $n \ge 2$, is a derivation.

math.RA↗
arXiv2016

Characterizations of centralizers and derivations on some algebras

A linear mapping $ϕ$ on an algebra $\mathcal{A}$ is called a centralizable mapping at $G\in\mathcal{A}$ if $ϕ(AB)=ϕ(A)B=Aϕ(B)$ for each $A$ and $B$ in $\mathcal{A}$ with $AB=G$, and $ϕ$ is called a derivable mapping at $G\in\mathcal{A}$ if $ϕ(AB)=ϕ(A)B+Aϕ(B)$ for each $A$ and $B$ in $\mathcal{A}$ with $AB=G$. A point $G$ in $\mathcal{A}$ is called a full-centralizable point (resp. full-derivable point) if every centralizable (resp. derivable) mapping at $G$ is a centralizer (resp. derivation). We prove that every point in a von Neumann algebra or a triangular algebra is a full-centralizable point. We also prove that a point in a von Neumann algebra is a full-derivable point if and only if its central carrier is the unit.

math.OA↗
arXiv2015

Similarity degree of a class of C$^*$-algebras

Suppose that $\mathcal M$ is a countably decomposable type II$_1$ von Neumann algebra and $\mathcal A$ is a separable, non-nuclear, unital C$^*$-algebra. We show that, if $\mathcal M$ has Property $Γ$, then the similarity degree of $\mathcal M$ is less than or equal to $5$. If $\mathcal A$ has Property c$^*$-$Γ$, then the similarity degree of $\mathcal A$ is equal to $3$. In particular, the similarity degree of a $\mathcal Z$-stable, separable, non-nuclear, unital C$^*$-algebra is equal to $3$.

math.OA↗
arXiv2014

Hochschild cohomology of type II$_1$ von Neumann algebras with Property $Γ$

In this paper, Property $Γ$ for a type II$_{1}$ von Neumann algebra is introduced as a generalization of Murray and von Neumann's Property $Γ$ for a type II$_{1}$ factor. The main result of this paper is that if a type II$_{1}$ von Neumann algebra $\mathcal{M}$ with separable predual has Property $Γ$, then the continuous Hochschild cohomology group $H^{k}(\mathcal{M}, \mathcal{M})$ vanishes for every $k \geq 2$. This gives a generalization of an earlier result due to E. Christensen, F. Pop, A.M. Sinclair and R.R. Smith.

math.OA↗