Pub Date : 2021-06-01DOI: 10.3724/sp.j.1089.2021.18574
Bai Jing, Wenhui Zhou, Jiwen Tuo, Feiwei Qin
: The existing sketch-based 3D model retrieval methods often regard data as static input, and utilize isting work, the accuracy rate is higher, which verifies the feasibility and effectiveness of the proposed algorithm.
{"title":"End-to-End Sketch-3D Model Retrieval with Spatiotemporal Information Joint Embedding","authors":"Bai Jing, Wenhui Zhou, Jiwen Tuo, Feiwei Qin","doi":"10.3724/sp.j.1089.2021.18574","DOIUrl":"https://doi.org/10.3724/sp.j.1089.2021.18574","url":null,"abstract":": The existing sketch-based 3D model retrieval methods often regard data as static input, and utilize isting work, the accuracy rate is higher, which verifies the feasibility and effectiveness of the proposed algorithm.","PeriodicalId":52442,"journal":{"name":"计算机辅助设计与图形学学报","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47852131","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}
Pub Date : 2021-06-01DOI: 10.3724/sp.j.1089.2021.18590
Naye Ji, Yong Gao, Youbing Zhao, Dingguo Yu, Shaowei Chu
Visual analysis of sports event news involves sports event analysis and news visualization, which plays an important role in quick generating of event news reports, enhancing the expression of event news, and assisting event analysis. A knowledge graph based visual analytics method is proposed for basketball data news generation. The interactive visualization tools are designed for basic matches and team data of basketball games, such as basketball player data, team data and spatio-temporal data, supplemented by knowledge graph, which enriches background knowledge. Besides the overview and statistical view of basic data, it also includes the visual design of knowledge map of the background data for the game and the visual view for the spatio-temporal data of the game, so that basketball fans and ordinary readers could benefit from a multi-dimensional perspective to enjoy the various aspects of the basketball game. Therefore, the visualization system can increase the richness of sports event news, and serve basketball fans and ordinary readers better. Through the evaluation by basketball fans, general readers, and professional sports journalists, most of the average scores of each visual module are above 4 points basically under a 5-point quantitative 838 计算机辅助设计与图形学学报 第 33 卷 scoring system. It is proved that the knowledge graph used in basketball event visualization has both vividness and knowledge, which enhances the expressiveness of basketball event news.
{"title":"Knowledge Graph Assisted Basketball Sport News Visualization","authors":"Naye Ji, Yong Gao, Youbing Zhao, Dingguo Yu, Shaowei Chu","doi":"10.3724/sp.j.1089.2021.18590","DOIUrl":"https://doi.org/10.3724/sp.j.1089.2021.18590","url":null,"abstract":"Visual analysis of sports event news involves sports event analysis and news visualization, which plays an important role in quick generating of event news reports, enhancing the expression of event news, and assisting event analysis. A knowledge graph based visual analytics method is proposed for basketball data news generation. The interactive visualization tools are designed for basic matches and team data of basketball games, such as basketball player data, team data and spatio-temporal data, supplemented by knowledge graph, which enriches background knowledge. Besides the overview and statistical view of basic data, it also includes the visual design of knowledge map of the background data for the game and the visual view for the spatio-temporal data of the game, so that basketball fans and ordinary readers could benefit from a multi-dimensional perspective to enjoy the various aspects of the basketball game. Therefore, the visualization system can increase the richness of sports event news, and serve basketball fans and ordinary readers better. Through the evaluation by basketball fans, general readers, and professional sports journalists, most of the average scores of each visual module are above 4 points basically under a 5-point quantitative 838 计算机辅助设计与图形学学报 第 33 卷 scoring system. It is proved that the knowledge graph used in basketball event visualization has both vividness and knowledge, which enhances the expressiveness of basketball event news.","PeriodicalId":52442,"journal":{"name":"计算机辅助设计与图形学学报","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46573692","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}
Pub Date : 2021-06-01DOI: 10.3724/sp.j.1089.2021.18537
Shuai Yuan, Kang Wang, Yi Shan, Jinfu Yang
: Existing object detection algorithms only use a fixed size convolution kernel when extracting features, ignoring the difference in the receptive field of different scale features, which affects the detection ef-fect of different scale objects. To solve this problem, a multi-scale object detection network based on multi-branch parallel dilated convolution is proposed. Firstly, the basic network VGG-16 is used to extract the features of the image. Secondly, a multi-branch parallel dilated convolution is designed to extract multi-scale features to improve object detection ability of the network. Then, a non-local block is employed to integrate the global spatial information and enhance the context information. Finally, the object detection and location tasks are performed on feature maps with different scales. Experimental results on PASCAL VOC and MS COCO datasets demonstrate that the proposed method can effectively improve the detection accuracy of different scale objects and clearly improve the detection accuracy of small objects.
{"title":"Multi-Scale Object Detection Method Based on Multi-Branch Parallel Dilated Convolution","authors":"Shuai Yuan, Kang Wang, Yi Shan, Jinfu Yang","doi":"10.3724/sp.j.1089.2021.18537","DOIUrl":"https://doi.org/10.3724/sp.j.1089.2021.18537","url":null,"abstract":": Existing object detection algorithms only use a fixed size convolution kernel when extracting features, ignoring the difference in the receptive field of different scale features, which affects the detection ef-fect of different scale objects. To solve this problem, a multi-scale object detection network based on multi-branch parallel dilated convolution is proposed. Firstly, the basic network VGG-16 is used to extract the features of the image. Secondly, a multi-branch parallel dilated convolution is designed to extract multi-scale features to improve object detection ability of the network. Then, a non-local block is employed to integrate the global spatial information and enhance the context information. Finally, the object detection and location tasks are performed on feature maps with different scales. Experimental results on PASCAL VOC and MS COCO datasets demonstrate that the proposed method can effectively improve the detection accuracy of different scale objects and clearly improve the detection accuracy of small objects.","PeriodicalId":52442,"journal":{"name":"计算机辅助设计与图形学学报","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44240523","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}
Pub Date : 2021-06-01DOI: 10.3724/sp.j.1089.2021.18589
Yuanchen Guo, Yun Cai, Songhai Zhang
Sketch-based image retrieval (SBIR) aims to return a collection of corresponding images based on an input sketch. Different from traditional content-based image retrieval, unique difficulties exist due to the large domain gap between sketches and natural images. Based on the similarity between edgemaps and sketches, a novel SBIR model named spatial attentive edgemap fusion is presented which combines both image and edgemap features. Images and the corresponding edgemaps are first encoded to their own latent feature space, and then fused by a learned spatial attention map. Experiment results on two widely-used SBIR datasets, Sketchy and Flickr15K, show the promising performance of the proposed model.
{"title":"Attentive Edgemap Fusion for Sketch-Based Image Retrieval","authors":"Yuanchen Guo, Yun Cai, Songhai Zhang","doi":"10.3724/sp.j.1089.2021.18589","DOIUrl":"https://doi.org/10.3724/sp.j.1089.2021.18589","url":null,"abstract":"Sketch-based image retrieval (SBIR) aims to return a collection of corresponding images based on an input sketch. Different from traditional content-based image retrieval, unique difficulties exist due to the large domain gap between sketches and natural images. Based on the similarity between edgemaps and sketches, a novel SBIR model named spatial attentive edgemap fusion is presented which combines both image and edgemap features. Images and the corresponding edgemaps are first encoded to their own latent feature space, and then fused by a learned spatial attention map. Experiment results on two widely-used SBIR datasets, Sketchy and Flickr15K, show the promising performance of the proposed model.","PeriodicalId":52442,"journal":{"name":"计算机辅助设计与图形学学报","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44111402","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}
Pub Date : 2021-06-01DOI: 10.3724/sp.j.1089.2021.18420
Shen Kecheng, Shi Quan, Wang Han
{"title":"Multimodal Visibility Deep Learning Model Based on Visible-Infrared Image Pair","authors":"Shen Kecheng, Shi Quan, Wang Han","doi":"10.3724/sp.j.1089.2021.18420","DOIUrl":"https://doi.org/10.3724/sp.j.1089.2021.18420","url":null,"abstract":"","PeriodicalId":52442,"journal":{"name":"计算机辅助设计与图形学学报","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47215645","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}
Pub Date : 2021-06-01DOI: 10.3724/sp.j.1089.2021.18639
Hui Wang, Ruizhen Hu
: In order to ensure the visual effect and unity of color mapping of multi-charts among the data with the same category, it is proposed to study the color assignment optimization problems on multi-charts, which introduces a score defined based on perception and harmony to guide the joint color assignment optimization. When given the data plotted with multi-charts and the color palette, first the quality of the color palette is evaluated. If the color palette is of high quality, the proposed method will directly optimize the color assignment. Otherwise, the color palette needs to be optimized first. Several user studies are conducted to show that the proposed method can ensure the unity of multi-charts color mapping and improve the quality of visualization results.
{"title":"Perception and Harmony Guided Color Assignment Optimization for Multi-Charts","authors":"Hui Wang, Ruizhen Hu","doi":"10.3724/sp.j.1089.2021.18639","DOIUrl":"https://doi.org/10.3724/sp.j.1089.2021.18639","url":null,"abstract":": In order to ensure the visual effect and unity of color mapping of multi-charts among the data with the same category, it is proposed to study the color assignment optimization problems on multi-charts, which introduces a score defined based on perception and harmony to guide the joint color assignment optimization. When given the data plotted with multi-charts and the color palette, first the quality of the color palette is evaluated. If the color palette is of high quality, the proposed method will directly optimize the color assignment. Otherwise, the color palette needs to be optimized first. Several user studies are conducted to show that the proposed method can ensure the unity of multi-charts color mapping and improve the quality of visualization results.","PeriodicalId":52442,"journal":{"name":"计算机辅助设计与图形学学报","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44373266","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}
Pub Date : 2021-06-01DOI: 10.3724/sp.j.1089.2021.18608
Zhe Zheng, Lin Lei, Hao Sun, Gangyao Kuang
: Remote sensing images of large scenes are complex, and have the characteristics of many catego-ries of objects, different scales and changeable directions, which lead to the problem of multi-class, multi-scale and multi-oriented of objects in remote sensing images. A remote sensing image object detection method based on multi-scale attention feature pyramid network (MAFPN) duce the redundant area in the bounding boxes, makes the predicted rotating bounding boxes fit the object more closely. The experimental results on the DOTA public dataset, compared with many classical detection algorithms based on convolutional neural networks, show that the average detection accuracy of the pro-posed method is significantly improved, which can detect objects of multi-scales and multi-oriented more accurately, and achieve the robust detection of multi-scale objects.
{"title":"FAGNet: Multi-Scale Object Detection Method in Remote Sensing Images by Combining MAFPN and GVR","authors":"Zhe Zheng, Lin Lei, Hao Sun, Gangyao Kuang","doi":"10.3724/sp.j.1089.2021.18608","DOIUrl":"https://doi.org/10.3724/sp.j.1089.2021.18608","url":null,"abstract":": Remote sensing images of large scenes are complex, and have the characteristics of many catego-ries of objects, different scales and changeable directions, which lead to the problem of multi-class, multi-scale and multi-oriented of objects in remote sensing images. A remote sensing image object detection method based on multi-scale attention feature pyramid network (MAFPN) duce the redundant area in the bounding boxes, makes the predicted rotating bounding boxes fit the object more closely. The experimental results on the DOTA public dataset, compared with many classical detection algorithms based on convolutional neural networks, show that the average detection accuracy of the pro-posed method is significantly improved, which can detect objects of multi-scales and multi-oriented more accurately, and achieve the robust detection of multi-scale objects.","PeriodicalId":52442,"journal":{"name":"计算机辅助设计与图形学学报","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45709965","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}
Pub Date : 2021-06-01DOI: 10.3724/sp.j.1089.2021.18563
Ziyao Meng, Sijia Chen, T. Lyu, Zhigang Zhang, Xiaoxia Wang, Bin Sheng, Lijuan Mao
The identification and classification of glomeruli in pathological sections is the key to diagnosing the degree and type of renal lesions. In order to solve the problem of glomerular recognition and classification, a complete glomerular detection and classification framework based on deep learning is designed. Glomeruli are detected and classified in the entire slice image. The framework includes four stages of glomerular recognition. In the first stage of scanning window generation, a new network framework, RGNet, is designed to initially deter948 计算机辅助设计与图形学学报 第 33 卷 mine the possible location of glomeruli. In the second stage of detection and coarse classification, Faster R-CNN is improved for glomerular data. In the third stage, the NMS-Lite algorithm is designed based on the NMS algorithm to merge the detected glomeruli. In the fourth stage of fine classification, two neural networks are trained using data augmentation to classify the degree of glomerular lesions. The experimental results has show that the glomerulus detection method proposed in this paper has achieved comparable accuracy on the test set with similar methods, and to a certain extent solves the problem that similar types of glomeruli are difficult to dis-
{"title":"Recognition and Classification of Glomerular Pathological Images Based on Deep Learning","authors":"Ziyao Meng, Sijia Chen, T. Lyu, Zhigang Zhang, Xiaoxia Wang, Bin Sheng, Lijuan Mao","doi":"10.3724/sp.j.1089.2021.18563","DOIUrl":"https://doi.org/10.3724/sp.j.1089.2021.18563","url":null,"abstract":"The identification and classification of glomeruli in pathological sections is the key to diagnosing the degree and type of renal lesions. In order to solve the problem of glomerular recognition and classification, a complete glomerular detection and classification framework based on deep learning is designed. Glomeruli are detected and classified in the entire slice image. The framework includes four stages of glomerular recognition. In the first stage of scanning window generation, a new network framework, RGNet, is designed to initially deter948 计算机辅助设计与图形学学报 第 33 卷 mine the possible location of glomeruli. In the second stage of detection and coarse classification, Faster R-CNN is improved for glomerular data. In the third stage, the NMS-Lite algorithm is designed based on the NMS algorithm to merge the detected glomeruli. In the fourth stage of fine classification, two neural networks are trained using data augmentation to classify the degree of glomerular lesions. The experimental results has show that the glomerulus detection method proposed in this paper has achieved comparable accuracy on the test set with similar methods, and to a certain extent solves the problem that similar types of glomeruli are difficult to dis-","PeriodicalId":52442,"journal":{"name":"计算机辅助设计与图形学学报","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2021-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47683790","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：