{"title":"Potential of Using Computer Vision to Predict Graphics for Learning-by-doing","authors":"Shaofu Li","doi":"10.1109/ICKII55100.2022.9983585","DOIUrl":null,"url":null,"abstract":"The learning and application of artificial intelligence (AI) is already a trend that higher education must deal with. Usually, schools train academic staff to become seed teachers and then deploy them in existing courses. Then, students have the opportunity to experience AI. We discuss the problems encountered by teachers in implementing blended teaching. For a case study of architectural learning, we investigate the discriminative ability of graphics, especially analytical graphics. The accuracy of such diagrams is often limited by the resolution of the mesh grid such as depthmapX for Space Syntax, which is well-known for the quantitative analysis of spatial relationships and social patterns in buildings and urban systems. We proposed the parameter settings of depthmapX, too. Judging from the initial application of Microsoft Lobe, the machine learning of vision has a higher error rate at low resolution. The result of this study is applied to the learning of computer vision and the discrimination and grade of students' homework.","PeriodicalId":352222,"journal":{"name":"2022 IEEE 5th International Conference on Knowledge Innovation and Invention (ICKII )","volume":"7 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 IEEE 5th International Conference on Knowledge Innovation and Invention (ICKII )","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICKII55100.2022.9983585","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The learning and application of artificial intelligence (AI) is already a trend that higher education must deal with. Usually, schools train academic staff to become seed teachers and then deploy them in existing courses. Then, students have the opportunity to experience AI. We discuss the problems encountered by teachers in implementing blended teaching. For a case study of architectural learning, we investigate the discriminative ability of graphics, especially analytical graphics. The accuracy of such diagrams is often limited by the resolution of the mesh grid such as depthmapX for Space Syntax, which is well-known for the quantitative analysis of spatial relationships and social patterns in buildings and urban systems. We proposed the parameter settings of depthmapX, too. Judging from the initial application of Microsoft Lobe, the machine learning of vision has a higher error rate at low resolution. The result of this study is applied to the learning of computer vision and the discrimination and grade of students' homework.