{"title":"基于相应三维纹理模型的三维人体重建:主要方法的比较","authors":"Hamid Habib Syed, M. H. Mahmood","doi":"10.1109/ICETECC56662.2022.10068940","DOIUrl":null,"url":null,"abstract":"It is the texture for materials development and the technology selected that determines the realism of the finished 3D model. An in-depth literature review is presented on approaches to producing textures for 3D models and reconstructing 3D humans. The process of texture mapping on a 3D model is one of the key steps. Making a model more realistic requires specific properties to be applied to the modeled object. Through reconstruction, insight into qualities of an object could be gained that aren’t apparent from a single perspective like its relative position and volume in relation to others in the scene. Human surface geometry is reconstructed, including invisible areas, and multiple image scales are combined into one 3D space. An extensive review of experimentation on datasets like Market-1501, DeepFashion and SURREAL shows that different analytical approaches work both qualitatively and quantitatively, as measured by evaluation metrics like SSIM and LPIPS. We aimed to estimate the current level of development and potential of these techniques so that there could be further improvements and a wider range of applications for texture generation and 3D human reconstruction.","PeriodicalId":364463,"journal":{"name":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2022-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"3D Human Reconstruction with Corresponding 3D Texture Model: A Comparison of Salient Approaches\",\"authors\":\"Hamid Habib Syed, M. H. Mahmood\",\"doi\":\"10.1109/ICETECC56662.2022.10068940\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"It is the texture for materials development and the technology selected that determines the realism of the finished 3D model. An in-depth literature review is presented on approaches to producing textures for 3D models and reconstructing 3D humans. The process of texture mapping on a 3D model is one of the key steps. Making a model more realistic requires specific properties to be applied to the modeled object. Through reconstruction, insight into qualities of an object could be gained that aren’t apparent from a single perspective like its relative position and volume in relation to others in the scene. Human surface geometry is reconstructed, including invisible areas, and multiple image scales are combined into one 3D space. An extensive review of experimentation on datasets like Market-1501, DeepFashion and SURREAL shows that different analytical approaches work both qualitatively and quantitatively, as measured by evaluation metrics like SSIM and LPIPS. We aimed to estimate the current level of development and potential of these techniques so that there could be further improvements and a wider range of applications for texture generation and 3D human reconstruction.\",\"PeriodicalId\":364463,\"journal\":{\"name\":\"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2022-12-07\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICETECC56662.2022.10068940\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 International Conference on Emerging Technologies in Electronics, Computing and Communication (ICETECC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICETECC56662.2022.10068940","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
3D Human Reconstruction with Corresponding 3D Texture Model: A Comparison of Salient Approaches
It is the texture for materials development and the technology selected that determines the realism of the finished 3D model. An in-depth literature review is presented on approaches to producing textures for 3D models and reconstructing 3D humans. The process of texture mapping on a 3D model is one of the key steps. Making a model more realistic requires specific properties to be applied to the modeled object. Through reconstruction, insight into qualities of an object could be gained that aren’t apparent from a single perspective like its relative position and volume in relation to others in the scene. Human surface geometry is reconstructed, including invisible areas, and multiple image scales are combined into one 3D space. An extensive review of experimentation on datasets like Market-1501, DeepFashion and SURREAL shows that different analytical approaches work both qualitatively and quantitatively, as measured by evaluation metrics like SSIM and LPIPS. We aimed to estimate the current level of development and potential of these techniques so that there could be further improvements and a wider range of applications for texture generation and 3D human reconstruction.