Chinese Conference on Pattern Recognition and Computer Vision最新文献

Towards Balanced RGB-TSDF Fusion for Consistent Semantic Scene Completion by 3D RGB Feature Completion and a Classwise Entropy Loss Function 通过 3D RGB 特征补全和分类熵损失函数实现均衡的 RGB-TSDF 融合以实现一致的语义场景补全

Chinese Conference on Pattern Recognition and Computer Vision Pub Date : 2024-03-25 DOI: 10.1007/978-981-99-8432-9_11

Laiyan Ding, Panwen Hu, Jie Li, Rui Huang

引用次数: 0

A Transformer-Based Adaptive Semantic Aggregation Method for UAV Visual Geo-Localization 基于变换器的无人机视觉地理定位自适应语义聚合方法

Chinese Conference on Pattern Recognition and Computer Vision Pub Date : 2024-01-03 DOI: 10.1007/978-981-99-8462-6_38

Shishen Li, Cuiwei Liu, Huaijun Qiu, Zhaokui Li

引用次数: 0

Few-Shot Segmentation via Rich Prototype Generation and Recurrent Prediction Enhancement 基于丰富原型生成和循环预测增强的少镜头分割

Chinese Conference on Pattern Recognition and Computer Vision Pub Date : 2022-10-03 DOI: 10.48550/arXiv.2210.00765

Hongsheng Wang, Xiaoqi Zhao, Youwei Pang, Jinqing Qi

{"title":"Few-Shot Segmentation via Rich Prototype Generation and Recurrent Prediction Enhancement","authors":"Hongsheng Wang, Xiaoqi Zhao, Youwei Pang, Jinqing Qi","doi":"10.48550/arXiv.2210.00765","DOIUrl":"https://doi.org/10.48550/arXiv.2210.00765","url":null,"abstract":". Prototype learning and decoder construction are the keys for few-shot segmentation. However, existing methods use only a single prototype generation mode, which can not cope with the intractable problem of objects with various scales. Moreover, the one-way forward propagation adopted by previous methods may cause information dilution from registered features during the decoding process. In this research, we propose a rich prototype generation module (RPGM) and a recurrent prediction enhancement module (RPEM) to reinforce the prototype learning paradigm and build a uniﬁed memory-augmented decoder for few-shot segmentation, respectively. Speciﬁcally, the RPGM combines superpixel and K-means clustering to generate rich prototype features with complementary scale relationships and adapt the scale gap between support and query images. The RPEM utilizes the recurrent mechanism to design a round-way propagation decoder. In this way, registered features can provide object-aware information continuously. Experiments show that our method consistently outperforms other competitors on two popular benchmarks PASCAL-5 i and COCO-20 i . role in few-shot segmentation. The prototype represents only object-related features and does not contain","PeriodicalId":420492,"journal":{"name":"Chinese Conference on Pattern Recognition and Computer Vision","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125502650","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

JVLDLoc: a Joint Optimization of Visual-LiDAR Constraints and Direction Priors for Localization in Driving Scenario JVLDLoc:基于视觉-激光雷达约束和方向先验的驾驶场景定位联合优化

Chinese Conference on Pattern Recognition and Computer Vision Pub Date : 2022-08-21 DOI: 10.48550/arXiv.2208.09777

Longrui Dong, Gang Zeng

引用次数: 0

Multiscale Autoencoder with Structural-Functional Attention Network for Alzheimer's Disease Prediction 基于结构-功能注意网络的多尺度自编码器用于阿尔茨海默病预测

Chinese Conference on Pattern Recognition and Computer Vision Pub Date : 2022-08-09 DOI: 10.1007/978-3-031-18910-4_24

Yongcheng Zong, Changhong Jing, Qiankun Zuo

引用次数: 3

Adversarial Learning Based Structural Brain-network Generative Model for Analyzing Mild Cognitive Impairment 基于对抗学习的结构脑网络生成模型分析轻度认知障碍

Chinese Conference on Pattern Recognition and Computer Vision Pub Date : 2022-08-09 DOI: 10.48550/arXiv.2208.08896

Heng Kong, Shuqiang Wang

{"title":"Adversarial Learning Based Structural Brain-network Generative Model for Analyzing Mild Cognitive Impairment","authors":"Heng Kong, Shuqiang Wang","doi":"10.48550/arXiv.2208.08896","DOIUrl":"https://doi.org/10.48550/arXiv.2208.08896","url":null,"abstract":"Mild cognitive impairment(MCI) is a precursor of Alzheimer’s disease(AD), and the detection of MCI is of great clinical significance. Analyzing the structural brain networks of patients is vital for the recognition of MCI. However, the current studies on structural brain networks are totally dependent on specific toolboxes, which is time-consuming and subjective. Few tools can obtain the structural brain networks from brain diffusion tensor images. In this work, an adversarial learning-based structural brain-network generative model(SBGM) is proposed to directly learn the structural connections from brain diffusion tensor images. By analyzing the differences in structural brain networks across subjects, we found that the structural brain networks of subjects showed a consistent trend from elderly normal controls(NC) to early mild cognitive impairment(EMCI) to late mild cognitive impairment(LMCI): structural connectivity progressed in a progressively weaker direction as the condition worsened. In addition, our proposed model tri-classifies EMCI, LMCI, and NC subjects, achieving a classification accuracy of 83.33% on the Alzheimer’s Disease Neuroimaging Initiative(ADNI) database.","PeriodicalId":420492,"journal":{"name":"Chinese Conference on Pattern Recognition and Computer Vision","volume":"7 5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130833117","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

SuperVessel: Segmenting High-resolution Vessel from Low-resolution Retinal Image 超级血管:从低分辨率视网膜图像中分割高分辨率血管

Chinese Conference on Pattern Recognition and Computer Vision Pub Date : 2022-07-28 DOI: 10.48550/arXiv.2207.13882

Yan Hu, Zhongxi Qiu, Dan Zeng, Li Jiang, Chen Lin, Jiang Liu

{"title":"SuperVessel: Segmenting High-resolution Vessel from Low-resolution Retinal Image","authors":"Yan Hu, Zhongxi Qiu, Dan Zeng, Li Jiang, Chen Lin, Jiang Liu","doi":"10.48550/arXiv.2207.13882","DOIUrl":"https://doi.org/10.48550/arXiv.2207.13882","url":null,"abstract":"Vascular segmentation extracts blood vessels from images and serves as the basis for diagnosing various diseases, like ophthalmic diseases. Ophthalmologists often require high-resolution segmentation results for analysis, which leads to super-computational load by most existing methods. If based on low-resolution input, they easily ignore tiny vessels or cause discontinuity of segmented vessels. To solve these problems, the paper proposes an algorithm named SuperVessel, which gives out high-resolution and accurate vessel segmentation using low-resolution images as input. We first take super-resolution as our auxiliary branch to provide potential high-resolution detail features, which can be deleted in the test phase. Secondly, we propose two modules to enhance the features of the interested segmentation region, including an upsampling with feature decomposition (UFD) module and a feature interaction module (FIM) with a constraining loss to focus on the interested features. Extensive experiments on three publicly available datasets demonstrate that our proposed SuperVessel can segment more tiny vessels with higher segmentation accuracy IoU over 6%, compared with other state-of-the-art algorithms. Besides, the stability of SuperVessel is also stronger than other algorithms. We will release the code after the paper is published.","PeriodicalId":420492,"journal":{"name":"Chinese Conference on Pattern Recognition and Computer Vision","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130437184","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 4

A high-order tensor completion algorithm based on Fully-Connected Tensor Network weighted optimization 基于全连通张量网络加权优化的高阶张量补全算法

Chinese Conference on Pattern Recognition and Computer Vision Pub Date : 2022-04-04 DOI: 10.48550/arXiv.2204.01732

Pei-Yi Yang, Yonghui Huang, Yuning Qiu, Weijun Sun, Guoxu Zhou

{"title":"A high-order tensor completion algorithm based on Fully-Connected Tensor Network weighted optimization","authors":"Pei-Yi Yang, Yonghui Huang, Yuning Qiu, Weijun Sun, Guoxu Zhou","doi":"10.48550/arXiv.2204.01732","DOIUrl":"https://doi.org/10.48550/arXiv.2204.01732","url":null,"abstract":". Tensor completion aimes at recovering missing data, and it is one of the popular concerns in deep learning and signal processing. Among the higher-order tensor decomposition algorithms, the recently proposed fully-connected tensor network decomposition (FCTN) algorithm is the most advanced. In this paper, by leveraging the superior expression of the fully-connected tensor network (FCTN) decomposition, we propose a new tensor completion method named the fully connected tensor network weighted optization(FCTN-WOPT). The algorithm per-forms a composition of the completed tensor by initialising the factors from the FCTN decomposition. We build a loss function with the weight tensor, the completed tensor and the incomplete tensor together, and then update the completed tensor using the lbfgs gradient descent algorithm to reduce the spatial memory occupation and speed up iterations. Finally we test the completion with synthetic data and real data (both image data and video data) and the results show the advanced performance of our FCTN-WOPT when it is applied to higher-order tensor completion.","PeriodicalId":420492,"journal":{"name":"Chinese Conference on Pattern Recognition and Computer Vision","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123198250","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 3

Enhancing Transferability of Adversarial Examples with Spatial Momentum 利用空间动量增强对抗性例子的可转移性

Chinese Conference on Pattern Recognition and Computer Vision Pub Date : 2022-03-25 DOI: 10.1007/978-3-031-18907-4_46

Guoqiu Wang, Huanqian Yan, Xingxing Wei

引用次数: 4

CLIP Meets Video Captioning: Concept-Aware Representation Learning Does Matter 剪辑与视频字幕:概念感知表示学习很重要

Chinese Conference on Pattern Recognition and Computer Vision Pub Date : 2021-11-30 DOI: 10.1007/978-3-031-18907-4_29

Bang Yang, Tong Zhang, Yuexian Zou

引用次数: 11