Computers & Graphics-Uk最新文献_第9页

Reconstruction-based distillation for anomaly detection 基于重构的异常检测蒸馏

IF 2.8 4区计算机科学

Computers & Graphics-Uk Pub Date : 2025-08-19 DOI: 10.1016/j.cag.2025.104328

Mengyang Zhao, Qiang Guo

{"title":"Reconstruction-based distillation for anomaly detection","authors":"Mengyang Zhao, Qiang Guo","doi":"10.1016/j.cag.2025.104328","DOIUrl":"10.1016/j.cag.2025.104328","url":null,"abstract":"<div><div>Anomaly detection plays an important role in industrial production. Recent advances have established knowledge distillation as a prominent anomaly detection method, leveraging the paradigm where a student network learns feature representations from a pre-trained teacher network. In practice, this traditional feature imitation strategy leads to overgeneralization, which degrades detection performance. To mitigate this limitation, this paper proposes a reconstruction-based distillation network that replaces direct feature imitation with feature reconstruction. This method improves the student network’s understanding of the semantic information of features. In addition, to improve the accuracy of the student network in predicting anomalous regions, we introduce a prediction consistency loss to ensure that the predictions of the student network are consistent in the training phase with the inference phase. Extensive experiments on the MVTec AD and VisA datasets validate the effectiveness and generalization capability of our method. On the MVTec AD benchmark, our method achieves 99.61% image-level AUROC for anomaly detection and 98.23% pixel-level AUROC for anomaly localization.</div></div>","PeriodicalId":50628,"journal":{"name":"Computers & Graphics-Uk","volume":"132 ","pages":"Article 104328"},"PeriodicalIF":2.8,"publicationDate":"2025-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144887492","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

ℂ3-palette: Co-saliency based colorization for comparing categorical visualizations 调色板：用于比较分类可视化的基于共显着性的着色

IF 2.8 4区计算机科学

Computers & Graphics-Uk Pub Date : 2025-08-18 DOI: 10.1016/j.cag.2025.104379

Kecheng Lu , Xubin Chai , Yi Hou , Yunhai Wang

引用次数: 0

Multi-view isosurface similarity analysis for transfer function design in direct volume rendering 直接体绘制中传递函数设计的多视点等值面相似性分析

IF 2.8 4区计算机科学

Computers & Graphics-Uk Pub Date : 2025-08-18 DOI: 10.1016/j.cag.2025.104343

Pei Li , Cheng Chen , Fang Wang , Xiaorong Zhang , Yaobin Wang

引用次数: 0

LiDAR-3DGS: LiDAR reinforcement for multimodal initialization of 3D Gaussian Splats LiDAR- 3dgs：用于三维高斯条纹多模态初始化的LiDAR增强

IF 2.8 4区计算机科学

Computers & Graphics-Uk Pub Date : 2025-08-16 DOI: 10.1016/j.cag.2025.104293

Hansol Lim , Hanbeom Chang , Jongseong Brad Choi , Chul Min Yeum

{"title":"LiDAR-3DGS: LiDAR reinforcement for multimodal initialization of 3D Gaussian Splats","authors":"Hansol Lim , Hanbeom Chang , Jongseong Brad Choi , Chul Min Yeum","doi":"10.1016/j.cag.2025.104293","DOIUrl":"10.1016/j.cag.2025.104293","url":null,"abstract":"<div><div>In this paper, we introduce LiDAR-3DGS, a novel approach for integrating LiDAR data into 3D Gaussian Splatting to enhance scene reconstructions. Rather than relying solely on image-based features, we integrate LiDAR-based features as initialization. To achieve this, we present a novel sampling technique – ChromaFilter – which prioritizes LiDAR points based on color diversity. It effectively samples important features while sparsifying redundant points. Experimental results on both a custom dataset and the ETH3D dataset show consistent improvements in PSNR and SSIM. A ChromaFilter sampling density of <em>n</em> = 10 yields a notable 7.064% gain on PSNR and 0.564% gain on SSIM on the custom dataset, while ETH3D reconstructions exhibit an average PSNR increase of 4.915% and SSIM gain of 0.5951%. Our method provides practical solution for incorporating LiDAR data into 3DGS. Because many operational industrial robots are already equipped with both LiDAR and cameras, our method can be easily adopted to industrial robots to reconstruct more accurate 3DGS models for engineering and remote inspections.</div></div>","PeriodicalId":50628,"journal":{"name":"Computers & Graphics-Uk","volume":"132 ","pages":"Article 104293"},"PeriodicalIF":2.8,"publicationDate":"2025-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144891865","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

FanNet: A mesh convolution operator for learning dense maps FanNet：用于学习密集映射的网格卷积算子

IF 2.8 4区计算机科学

Computers & Graphics-Uk Pub Date : 2025-08-16 DOI: 10.1016/j.cag.2025.104320

Güneş Sucu, Sinan Kalkan, Yusuf Sahillioğlu

引用次数: 0

Neural implicit curve: A robust curve modeling approach on surface meshes 神经隐式曲线：曲面网格的鲁棒曲线建模方法

IF 2.8 4区计算机科学

Computers & Graphics-Uk Pub Date : 2025-08-14 DOI: 10.1016/j.cag.2025.104351

Jintong Wang , Qi Zhang , Yun Zhang , Yao Jin , Huaxiong Zhang , Lili He

{"title":"Neural implicit curve: A robust curve modeling approach on surface meshes","authors":"Jintong Wang , Qi Zhang , Yun Zhang , Yao Jin , Huaxiong Zhang , Lili He","doi":"10.1016/j.cag.2025.104351","DOIUrl":"10.1016/j.cag.2025.104351","url":null,"abstract":"<div><div>Traditional implicit curve modeling methods on surface meshes, such as variational approaches, are often plagued by numerical instability and heavy reliance on mesh quality, severely limiting their reliability in practical applications. To address these challenges, we propose Neural Implicit Curve Modeling on Meshes (NICMM), a novel framework that integrates Neural Implicit Method with geometric constraints for robust curve design. NICMM leverages physics-driven loss functions to encode positional, smoothness, and other customized constraints, alleviates numerical instabilities and inaccuracies arising from low-quality meshes, such as convergence failures. The framework incorporates specialized modules (e.g., Efficient Channel Attention and Light GLU) to enhance feature extraction and computational efficiency and introduces a two-stage training strategy combining pre-training with rapid convergence optimization. Extensive experiments on the SHREC16 dataset demonstrate that NICMM has proven its mettle by outperforming traditional variational approaches in robustness. In the face of highly degraded meshes replete with elongated and near-degenerate elements, NICMM not only excels in generating high-fidelity curves but also maintains computational efficiency comparable to existing variational method, thereby showcasing its remarkable balance between accuracy and performance. Furthermore, NICMM also supports feature-aware curve design, enabling alignment with user-specified regions and obstacle avoidance through a unified guidance mechanism. This work establishes a new paradigm for manifold curve modeling, with significant potential in CAD/CAM systems, virtual surgery, and other domains that require precise and adaptive geometric design.</div></div>","PeriodicalId":50628,"journal":{"name":"Computers & Graphics-Uk","volume":"132 ","pages":"Article 104351"},"PeriodicalIF":2.8,"publicationDate":"2025-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144893095","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A generic query-modify framework for volumetric mesh processing 一种用于体积网格处理的通用查询修改框架

IF 2.8 4区计算机科学

Computers & Graphics-Uk Pub Date : 2025-08-14 DOI: 10.1016/j.cag.2025.104371

Guillaume Damiand , Vincent Nivoliers , Romain Pascual

引用次数: 0

Training-free style transfer via content-style image inversion 通过内容样式图像反转的无训练风格迁移

IF 2.8 4区计算机科学

Computers & Graphics-Uk Pub Date : 2025-08-14 DOI: 10.1016/j.cag.2025.104352

Songlin Lei , Qiuxia Yang , Ke Yang , Zhengpeng Zhao , Yuanyuan Pu

引用次数: 0

SDLKF: Signed Distance Linear Kernel Function for surface reconstruction SDLKF：曲面重构的有符号距离线性核函数