arXiv subjects

Qiyu Chen

Publications and source records attributed to Qiyu Chen.

At least 19 recordsLinked to original sources

arXiv2026

Task-Oriented Wave Processing with Stacked Intelligent Metasurfaces: Framework, Fusion, and Challenges

The deep integration of diverse services in sixth-generation (6G) networks poses significant challenges to conventional task-agnostic channels, often resulting in performance conflicts. To resolve these bottlenecks, this article introduces a physical-layer computing paradigm enabled by stacked intelligent metasurfaces (SIMs), transforming the wireless environment from a passive medium into a programmable signal processor. Specifically, we establish a unified framework to map high-level service requirements directly to wave-domain synthesis. We then investigate the fusion of diverse services, demonstrating how the deep computational architecture of SIMs resolves resource conflicts in integrated sensing and communication (ISAC) and integrated communication and computation (ICC) scenarios. Furthermore, we critically analyze fundamental challenges, including diffractive channel modeling and inverse task-to-phase mapping, while validating through numerical results that this approach elevates the system from simple coexistence to true service symbiosis. Finally, we discuss key research directions to pave the way for service-native 6G architectures.

cs.IT↗

arXiv2026

MRAD: Zero-Shot Anomaly Detection with Memory-Driven Retrieval

Zero-shot anomaly detection (ZSAD) often leverages pretrained vision or vision-language models, but many existing methods use prompt learning or complex modeling to fit the data distribution, resulting in high training or inference cost and limited cross-domain stability. To address these limitations, we propose Memory-Retrieval Anomaly Detection method (MRAD), a unified framework that replaces parametric fitting with a direct memory retrieval. The train-free base model, MRAD-TF, freezes the CLIP image encoder and constructs a two-level memory bank (image-level and pixel-level) from auxiliary data, where feature-label pairs are explicitly stored as keys and values. During inference, anomaly scores are obtained directly by similarity retrieval over the memory bank. Based on the MRAD-TF, we further propose two lightweight variants as enhancements: (i) MRAD-FT fine-tunes the retrieval metric with two linear layers to enhance the discriminability between normal and anomaly; (ii) MRAD-CLIP injects the normal and anomalous region priors from the MRAD-FT as dynamic biases into CLIP's learnable text prompts, strengthening generalization to unseen categories. Across 16 industrial and medical datasets, the MRAD framework consistently demonstrates superior performance in anomaly classification and segmentation, under both train-free and training-based settings. Our work shows that fully leveraging the empirical distribution of raw data, rather than relying only on model fitting, can achieve stronger anomaly detection performance. The code will be publicly released at https://github.com/CROVO1026/MRAD.

cs.CV↗

arXiv2025

Circular foliations and shear-radius coordinates on Teichm\"uller spaces of hyperbolic cone surfaces

We study the Teichm\"uller space $\mathcal{T}(S,\underline{p})$ of hyperbolic cone-surfaces of fixed topological type with marked cone singularities. Fix a combinatorial triangulation $G$, and let $\mathcal{T}(G)\subset \mathcal{T}(S,\underline{p})$ be the locus where $G$ admits a geodesic realization; varying $G$, these loci form an open cover of $\mathcal{T}(S,\underline{p})$. On $\mathcal{T}(G)$ we construct a circular foliation adapted to geodesic triangular complementary regions, which is naturally decomposed into interior and peripheral parts. This decomposition defines shear parameters on edges and radius parameters at the singularities, and yields global coordinates on $\mathcal{T}(G)$: the resulting shear-radius map is a homeomorphism onto an explicit open cone in a finite-dimensional real vector space. In the spirit of Thurston, we then introduce partial stretch and anti-stretch deformations by rescaling the transverse measures of the interior or peripheral components. Peripheral stretch rays converge, in the simple-curve length-spectrum topology, to the cusped hyperbolic metric determined by the shear data, while interior anti-stretch rays converge to a circle-packed hyperbolic cone metric determined by the radii. Finally, we give criteria for the realization of prescribed cone angles for fixed $G$ and prove sharp upper bounds for admissible cone angles on the universally triangulable locus.

math.GT↗

arXiv2025

DeltaDeno: Zero-Shot Anomaly Generation via Delta-Denoising Attribution

Anomaly generation is often framed as few-shot fine-tuning with anomalous samples, which contradicts the scarcity that motivates generation and tends to overfit category priors. We tackle the setting where no real anomaly samples or training are available. We propose Delta-Denoising (\textbf{DeltaDeno}), a training-free zero-shot anomaly generation method that localizes and edits defects by contrasting two diffusion branches driven by a minimal prompt pair under a shared schedule. By accumulating per-step denoising deltas into an image-specific localization map, we obtain a mask to guide the latent inpainting during later diffusion steps and preserve the surrounding context while generating realistic local defects. To improve stability and control, DeltaDeno performs token-level prompt refinement that aligns shared content and strengthens anomaly tokens, and applies a spatial attention bias restricted to anomaly tokens in the predicted region. Experiments on public datasets show that DeltaDeno achieves great generation, realism and consistent gains in downstream detection performance. Code will be made publicly available at https://github.com/CROVO1026/DeltaDeno.

cs.CV↗

arXiv2025

Dexterous Manipulation Transfer via Progressive Kinematic-Dynamic Alignment

The inherent difficulty and limited scalability of collecting manipulation data using multi-fingered robot hand hardware platforms have resulted in severe data scarcity, impeding research on data-driven dexterous manipulation policy learning. To address this challenge, we present a hand-agnostic manipulation transfer system. It efficiently converts human hand manipulation sequences from demonstration videos into high-quality dexterous manipulation trajectories without requirements of massive training data. To tackle the multi-dimensional disparities between human hands and dexterous hands, as well as the challenges posed by high-degree-of-freedom coordinated control of dexterous hands, we design a progressive transfer framework: first, we establish primary control signals for dexterous hands based on kinematic matching; subsequently, we train residual policies with action space rescaling and thumb-guided initialization to dynamically optimize contact interactions under unified rewards; finally, we compute wrist control trajectories with the objective of preserving operational semantics. Using only human hand manipulation videos, our system automatically configures system parameters for different tasks, balancing kinematic matching and dynamic optimization across dexterous hands, object categories, and tasks. Extensive experimental results demonstrate that our framework can automatically generate smooth and semantically correct dexterous hand manipulation that faithfully reproduces human intentions, achieving high efficiency and strong generalizability with an average transfer success rate of 73%, providing an easily implementable and scalable method for collecting robot dexterous manipulation data.

cs.RO↗

arXiv2025

Anomagic: Crossmodal Prompt-driven Zero-shot Anomaly Generation

We propose Anomagic, a zero-shot anomaly generation method that produces semantically coherent anomalies without requiring any exemplar anomalies. By unifying both visual and textual cues through a crossmodal prompt encoding scheme, Anomagic leverages rich contextual information to steer an inpainting-based generation pipeline. A subsequent contrastive refinement strategy enforces precise alignment between synthesized anomalies and their masks, thereby bolstering downstream anomaly detection accuracy. To facilitate training, we introduce AnomVerse, a collection of 12,987 anomaly-mask-caption triplets assembled from 13 publicly available datasets, where captions are automatically generated by multimodal large language models using structured visual prompts and template-based textual hints. Extensive experiments demonstrate that Anomagic trained on AnomVerse can synthesize more realistic and varied anomalies than prior methods, yielding superior improvements in downstream anomaly detection. Furthermore, Anomagic can generate anomalies for any normal-category image using user-defined prompts, establishing a versatile foundation model for anomaly generation.

cs.CV↗

arXiv2025

Aligning Inductive Bias for Data-Efficient Generalization in State Space Models

The remarkable success of modern AI has been closely tied to scaling laws, yet the finite supply of high-quality data makes data efficiency--learning more from less--an increasingly important frontier. A model's inductive bias is a critical lever for data efficiency, but foundational sequence models such as State Space Models (SSMs) often rely on fixed, task-agnostic biases. When this fixed prior is misaligned with the underlying structure of a task, the model may require additional samples to overcome its own bias before learning the relevant signal. In this work, we introduce a principled framework for understanding and aligning the inductive bias of linear time-invariant SSMs. We first formalize this bias through an SSM-induced kernel and show theoretically and empirically that its spectrum is governed by the model's frequency response. This characterization motivates Task-Dependent Initialization (TDI), a fast power-spectrum matching method that aligns the initial SSM bias with the task's spectral characteristics before downstream training. Across controlled synthetic experiments, trainable one-layer SSMs, and deep SSMs on diverse real-world benchmarks, TDI can improve data-efficient generalization primarily when task-relevant spectral structure is present and the default SSM bias is spectrally mismatched. Our results provide both a theoretical lens and a practical tool for task-adaptive inductive bias, suggesting a path toward more data-efficient sequence modeling.

cs.LG↗

arXiv2025

CoPS: Conditional Prompt Synthesis for Zero-Shot Anomaly Detection

Recently, large pre-trained vision-language models have shown remarkable performance in zero-shot anomaly detection (ZSAD). With fine-tuning on a single auxiliary dataset, the model enables cross-category anomaly detection on diverse datasets covering industrial defects and medical lesions. Compared to manually designed prompts, prompt learning eliminates the need for expert knowledge and trial-and-error. However, it still faces the following challenges: (i) static learnable tokens struggle to capture the continuous and diverse patterns of normal and anomalous states, limiting generalization to unseen categories; (ii) fixed textual labels provide overly sparse category information, making the model prone to overfitting to a specific semantic subspace. To address these issues, we propose Conditional Prompt Synthesis (CoPS), a novel framework that synthesizes dynamic prompts conditioned on visual features to enhance ZSAD performance. Specifically, we extract representative normal and anomaly prototypes from fine-grained patch features and explicitly inject them into prompts, enabling adaptive state modeling. Given the sparsity of class labels, we leverage a variational autoencoder to model semantic image features and implicitly fuse varied class tokens into prompts. Additionally, integrated with our spatially-aware alignment mechanism, extensive experiments demonstrate that CoPS surpasses state-of-the-art methods by 1.4% in classification AUROC and 1.9% in segmentation AUROC across 13 industrial and medical datasets. The code is available at https://github.com/cqylunlun/CoPS.

cs.CV↗

arXiv2025

INP-Former++: Advancing Universal Anomaly Detection via Intrinsic Normal Prototypes and Residual Learning

Anomaly detection (AD) is essential for industrial inspection and medical diagnosis, yet existing methods typically rely on ``comparing'' test images to normal references from a training set. However, variations in appearance and positioning often complicate the alignment of these references with the test image, limiting detection accuracy. We observe that most anomalies manifest as local variations, meaning that even within anomalous images, valuable normal information remains. We argue that this information is useful and may be more aligned with the anomalies since both the anomalies and the normal information originate from the same image. Therefore, rather than relying on external normality from the training set, we propose INP-Former, a novel method that extracts Intrinsic Normal Prototypes (INPs) directly from the test image. Specifically, we introduce the INP Extractor, which linearly combines normal tokens to represent INPs. We further propose an INP Coherence Loss to ensure INPs can faithfully represent normality for the testing image. These INPs then guide the INP-guided Decoder to reconstruct only normal tokens, with reconstruction errors serving as anomaly scores. Additionally, we propose a Soft Mining Loss to prioritize hard-to-optimize samples during training. INP-Former achieves state-of-the-art performance in single-class, multi-class, and few-shot AD tasks across MVTec-AD, VisA, and Real-IAD, positioning it as a versatile and universal solution for AD. Remarkably, INP-Former also demonstrates some zero-shot AD capability. Furthermore, we propose a soft version of the INP Coherence Loss and enhance INP-Former by incorporating residual learning, leading to the development of INP-Former++. The proposed method significantly improves detection performance across single-class, multi-class, semi-supervised, few-shot, and zero-shot settings.

cs.CV↗

arXiv2025

Center-aware Residual Anomaly Synthesis for Multi-class Industrial Anomaly Detection

Anomaly detection plays a vital role in the inspection of industrial images. Most existing methods require separate models for each category, resulting in multiplied deployment costs. This highlights the challenge of developing a unified model for multi-class anomaly detection. However, the significant increase in inter-class interference leads to severe missed detections. Furthermore, the intra-class overlap between normal and abnormal samples, particularly in synthesis-based methods, cannot be ignored and may lead to over-detection. To tackle these issues, we propose a novel Center-aware Residual Anomaly Synthesis (CRAS) method for multi-class anomaly detection. CRAS leverages center-aware residual learning to couple samples from different categories into a unified center, mitigating the effects of inter-class interference. To further reduce intra-class overlap, CRAS introduces distance-guided anomaly synthesis that adaptively adjusts noise variance based on normal data distribution. Experimental results on diverse datasets and real-world industrial applications demonstrate the superior detection accuracy and competitive inference speed of CRAS. The source code and the newly constructed dataset are publicly available at https://github.com/cqylunlun/CRAS.

cs.CV↗

arXiv2025

Bayesian Prompt Flow Learning for Zero-Shot Anomaly Detection

Recently, vision-language models (e.g. CLIP) have demonstrated remarkable performance in zero-shot anomaly detection (ZSAD). By leveraging auxiliary data during training, these models can directly perform cross-category anomaly detection on target datasets, such as detecting defects on industrial product surfaces or identifying tumors in organ tissues. Existing approaches typically construct text prompts through either manual design or the optimization of learnable prompt vectors. However, these methods face several challenges: 1) handcrafted prompts require extensive expert knowledge and trial-and-error; 2) single-form learnable prompts struggle to capture complex anomaly semantics; and 3) an unconstrained prompt space limits generalization to unseen categories. To address these issues, we propose Bayesian Prompt Flow Learning (Bayes-PFL), which models the prompt space as a learnable probability distribution from a Bayesian perspective. Specifically, a prompt flow module is designed to learn both image-specific and image-agnostic distributions, which are jointly utilized to regularize the text prompt space and improve the model's generalization on unseen categories. These learned distributions are then sampled to generate diverse text prompts, effectively covering the prompt space. Additionally, a residual cross-model attention (RCA) module is introduced to better align dynamic text embeddings with fine-grained image features. Extensive experiments on 15 industrial and medical datasets demonstrate our method's superior performance. The code is available at https://github.com/xiaozhen228/Bayes-PFL.

cs.CV↗

arXiv2024

Progressive Boundary Guided Anomaly Synthesis for Industrial Anomaly Detection

Unsupervised anomaly detection methods can identify surface defects in industrial images by leveraging only normal samples for training. Due to the risk of overfitting when learning from a single class, anomaly synthesis strategies are introduced to enhance detection capability by generating artificial anomalies. However, existing strategies heavily rely on anomalous textures from auxiliary datasets. Moreover, their limitations in the coverage and directionality of anomaly synthesis may result in a failure to capture useful information and lead to significant redundancy. To address these issues, we propose a novel Progressive Boundary-guided Anomaly Synthesis (PBAS) strategy, which can directionally synthesize crucial feature-level anomalies without auxiliary textures. It consists of three core components: Approximate Boundary Learning (ABL), Anomaly Feature Synthesis (AFS), and Refined Boundary Optimization (RBO). To make the distribution of normal samples more compact, ABL first learns an approximate decision boundary by center constraint, which improves the center initialization through feature alignment. AFS then directionally synthesizes anomalies with more flexible scales guided by the hypersphere distribution of normal features. Since the boundary is so loose that it may contain real anomalies, RBO refines the decision boundary through the binary classification of artificial anomalies and normal features. Experimental results show that our method achieves state-of-the-art performance and the fastest detection speed on three widely used industrial datasets, including MVTec AD, VisA, and MPDD. The code will be available at: https://github.com/cqylunlun/PBAS.

cs.CV↗

arXiv2024

A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and Localization

Anomaly synthesis strategies can effectively enhance unsupervised anomaly detection. However, existing strategies have limitations in the coverage and controllability of anomaly synthesis, particularly for weak defects that are very similar to normal regions. In this paper, we propose Global and Local Anomaly co-Synthesis Strategy (GLASS), a novel unified framework designed to synthesize a broader coverage of anomalies under the manifold and hypersphere distribution constraints of Global Anomaly Synthesis (GAS) at the feature level and Local Anomaly Synthesis (LAS) at the image level. Our method synthesizes near-in-distribution anomalies in a controllable way using Gaussian noise guided by gradient ascent and truncated projection. GLASS achieves state-of-the-art results on the MVTec AD (detection AUROC of 99.9\%), VisA, and MPDD datasets and excels in weak defect detection. The effectiveness and efficiency have been further validated in industrial applications for woven fabric defect detection. The code and dataset are available at: \url{https://github.com/cqylunlun/GLASS}.

cs.CV↗

arXiv2024

A Unified Search and Recommendation Framework Based on Multi-Scenario Learning for Ranking in E-commerce

Search and recommendation (S&R) are the two most important scenarios in e-commerce. The majority of users typically interact with products in S&R scenarios, indicating the need and potential for joint modeling. Traditional multi-scenario models use shared parameters to learn the similarity of multiple tasks, and task-specific parameters to learn the divergence of individual tasks. This coarse-grained modeling approach does not effectively capture the differences between S&R scenarios. Furthermore, this approach does not sufficiently exploit the information across the global label space. These issues can result in the suboptimal performance of multi-scenario models in handling both S&R scenarios. To address these issues, we propose an effective and universal framework for Unified Search and Recommendation (USR), designed with S&R Views User Interest Extractor Layer (IE) and S&R Views Feature Generator Layer (FG) to separately generate user interests and scenario-agnostic feature representations for S&R. Next, we introduce a Global Label Space Multi-Task Layer (GLMT) that uses global labels as supervised signals of auxiliary tasks and jointly models the main task and auxiliary tasks using conditional probability. Extensive experimental evaluations on real-world industrial datasets show that USR can be applied to various multi-scenario models and significantly improve their performance. Online A/B testing also indicates substantial performance gains across multiple metrics. Currently, USR has been successfully deployed in the 7Fresh App.

cs.IR↗

arXiv2024

Convex surfaces with prescribed induced metrics in anti-de Sitter spacetimes

Let $S$ be a closed surface of genus at least $2$, let $h$ be a smooth metric of curvature $K<-1$ on $S$, and let $h_0$ be a hyperbolic metric on $S$. We show that there exists a unique quasifuchsian AdS spacetime with left metric isotopic to $h_0$, containing a past-convex Cauchy surface with induced metric isotopic to $h$.

math.DG↗

arXiv2024

The geometric data on the boundary of convex subsets of hyperbolic manifolds

Let $N$ be a geodesically convex subset in a convex co-compact hyperbolic manifold $M$ with incompressible boundary. We assume that each boundary component of $N$ is either a boundary component of $\partial_\infty M$, or a smooth, locally convex surface in $M$. We show that $N$ is uniquely determined by the boundary data defined by the conformal structure on the boundary components at infinity, and by either the induced metric or the third fundamental form on the boundary components which are locally convex surfaces. We also describe the possible boundary data. This provides an extension of both the hyperbolic Weyl problem and the Ahlfors-Bers Theorem. Using this statement for quasifuchsian manifolds, we obtain existence results for similar questions for convex domains $Ω\subset \HH^3$ which meets the boundary at infinity $\partial_{\infty}\HH^3$ either along a quasicircle or along a quasidisk. The boundary data then includes either the induced metric or the third fundamental form in $\HH^3$, but also an additional "gluing" data between different components of the boundary, either in $\HH^3$ or in $\partial_\infty\HH^3$.

math.DG↗

arXiv2024

Unveiling microstructural damage for leakage current degradation in SiC Schottky diode after heavy ions irradiation under 200 V

Single-event burnout and single-event leakage current (SELC) in SiC power devices induced by heavy ions severely limit their space application, and the underlying mechanism is still unclear. One fundamental problem is lack of high-resolution characterization of radiation damage in the irradiated SiC power devices, which is a crucial indicator of the related mechanism. In this letter, high-resolution transmission electron microscopy (TEM) was used to characterize the radiation damage in the 1437.6 MeV 181Ta-irradiated SiC junction barrier Schottky diode under 200 V. The amorphous radiation damage with about 52 nm in diameter and 121 nm in length at the Schottky metal (Ti)-semiconductor (SiC) interface was observed. More importantly, in the damage site the atomic mixing of Ti, Si, and C was identified by electron energy loss spectroscopy and high-angle annular dark-field scanning TEM. It indicates that the melting of the Ti-SiC interface induced by localized Joule heating is responsible for the amorphization and the formation of titanium silicide, titanium carbide, or ternary phases. These modifications at nanoscale in turn cause the localized degradation of the Schottky contact, resulting in the permanent increase in leakage current. This experimental study provides very valuable clues to thorough understanding of the SELC mechanism in SiC diode.

cond-mat.mtrl-sci↗

arXiv2021

Hyperideal polyhedra in the 3-dimensional anti-de Sitter space

We study hyperideal polyhedra in the 3-dimensional anti-de Sitter space $AdS^3$, which are defined as the intersection of the projective model of $AdS^3$ with a convex polyhedron in $RP^3$ whose vertices are all outside of $AdS^3$ and whose edges all meet $AdS^3$. We show that hyperideal polyhedra in $AdS^3$ are uniquely determined by their combinatorics and dihedral angles, as well as by the induced metric on their boundary together with an additional combinatorial data, and describe the possible dihedral angles and the possible induced metrics on the boundary.

math.DG↗