The use of self-avatars in a VR application can enhance presence and embodiment which leads to a better user experience. In collaborative VR it also facilitates non-verbal communication. Currently it is possible to track a few body parts with cheap trackers and then apply IK methods to animate a character. However, the correspondence between trackers and avatar joints is typically fixed ad-hoc, which is enough to animate the avatar, but causes noticeable mismatches between the user's body pose and the avatar. In this paper we present a fast and easy to set up system to compute exact offset values, unique for each user, which leads to improvements in avatar movement. Our user study shows that the Sense of Embodiment increased significantly when using exact offsets as opposed to fixed ones. We also allowed the users to see a semitransparent avatar overlaid with their real body to objectively evaluate the quality of the avatar movement with our technique.
{"title":"AvatarGo: Plug and Play self-avatars for VR","authors":"J. L. Ponton, E. Monclús, N. Pelechano","doi":"10.2312/egs.20221037","DOIUrl":"https://doi.org/10.2312/egs.20221037","url":null,"abstract":"The use of self-avatars in a VR application can enhance presence and embodiment which leads to a better user experience. In collaborative VR it also facilitates non-verbal communication. Currently it is possible to track a few body parts with cheap trackers and then apply IK methods to animate a character. However, the correspondence between trackers and avatar joints is typically fixed ad-hoc, which is enough to animate the avatar, but causes noticeable mismatches between the user's body pose and the avatar. In this paper we present a fast and easy to set up system to compute exact offset values, unique for each user, which leads to improvements in avatar movement. Our user study shows that the Sense of Embodiment increased significantly when using exact offsets as opposed to fixed ones. We also allowed the users to see a semitransparent avatar overlaid with their real body to objectively evaluate the quality of the avatar movement with our technique.","PeriodicalId":72958,"journal":{"name":"Eurographics ... Workshop on 3D Object Retrieval : EG 3DOR. Eurographics Workshop on 3D Object Retrieval","volume":"26 1","pages":"77-80"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87928444","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}
E Sariyanidi, C J Zampella, M N Drye, M L Fecher, G Megginson, L Soskey Cubit, R T Schultz, W Guthrie, B Tunc
3D morphable models (3DMMs) simultaneously reconstruct facial morphology, expression and pose from 2D images, and thus could be an invaluable tool for capturing and characterizing the face and facial behavior in early childhood. However, 3DMM fitting on infants is a largely unexplored problem. All publicly available 3DMMs are developed for adults, and it is unclear if and to what extent they can be used on videos of infants. In this paper, we compare five state-of-the-art 3DMM fitting methods on data from naturalistic infant-caregiver interactions. Results suggest that it is possible to produce consistent and subject-specific reconstructions of 3D shape identity from multiple frames, but not from a single frame. Qualitative evaluation highlights that facial regions with high texture variation, such as eyes, brows and mouth, are captured with higher accuracy compared to the rest of the face. Thus, even though a 3DMM developed for adults has significant limitations when reconstructing the morphology of the entire facial region of infants, applications that involve analysis of facial behavior can be feasible. Our encouraging results, combined with the unique ability of 3DMMs to disentangle two major sources of noise for expression analysis (i.e., identity bias and pose variations), motivate future research on using 3DMMs to measure the facial behavior of infants.
{"title":"Reconstructing 3D Face of Infants in Social Interactions Using Morphable Models of Non-Infants.","authors":"E Sariyanidi, C J Zampella, M N Drye, M L Fecher, G Megginson, L Soskey Cubit, R T Schultz, W Guthrie, B Tunc","doi":"10.2312/3dor.20221178","DOIUrl":"10.2312/3dor.20221178","url":null,"abstract":"<p><p>3D morphable models (3DMMs) simultaneously reconstruct facial morphology, expression and pose from 2D images, and thus could be an invaluable tool for capturing and characterizing the face and facial behavior in early childhood. However, 3DMM fitting on infants is a largely unexplored problem. All publicly available 3DMMs are developed for adults, and it is unclear if and to what extent they can be used on videos of infants. In this paper, we compare five state-of-the-art 3DMM fitting methods on data from naturalistic infant-caregiver interactions. Results suggest that it is possible to produce consistent and subject-specific reconstructions of 3D shape identity from multiple frames, but not from a single frame. Qualitative evaluation highlights that facial regions with high texture variation, such as eyes, brows and mouth, are captured with higher accuracy compared to the rest of the face. Thus, even though a 3DMM developed for adults has significant limitations when reconstructing the morphology of the entire facial region of infants, applications that involve analysis of facial behavior can be feasible. Our encouraging results, combined with the unique ability of 3DMMs to disentangle two major sources of noise for expression analysis (i.e., identity bias and pose variations), motivate future research on using 3DMMs to measure the facial behavior of infants.</p>","PeriodicalId":72958,"journal":{"name":"Eurographics ... Workshop on 3D Object Retrieval : EG 3DOR. Eurographics Workshop on 3D Object Retrieval","volume":"2022 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9645458/pdf/nihms-1836700.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"9150797","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-27DOI: 10.14324/111.444.2396-9008.1340
Glynnis Stevenson
This review looks at the Kemper Museum of Contemporary Art’s tribute to the past decade of artwork by Dyani White Hawk, a curator and multimedia artist of Sičáŋǧu Lakota, German and Welsh ancestry. The exhibition entrance sits directly across from Sean Scully’s The Moroccan (1984), positioning White Hawk’s work as a commentary on the Euro-American appropriation of non-white aesthetics. See for example, her canvas Untitled (coral, turquoise, and yellow) (2016), which embeds intricate rows of blue and yellow beads within bands of orange and red paint that evoke colour field painting. White Hawk receives pride of place in the museum’s main exhibition space, a series of four rooms showcasing the artist’s skill in many media: acrylic painting, beadwork, video, and photography. She layers and leverages Indigenous motifs amidst those associated with Euro-American artists. The exhibition tackles the debate over the false binary of art vs. craft and the centrality of indigenous women to Native American communities; the exhibition tackles the former deftly, the latter feels less fully explored.
这篇评论着眼于肯珀当代艺术博物馆(Kemper Museum of Contemporary Art)对Dyani White Hawk过去十年作品的致敬,Dyani White Hawk是一位策展人和多媒体艺术家,拥有Sičáŋǧu拉科塔、德国和威尔士血统。展览入口正对面是Sean Scully的《摩洛哥人》(1984),将白鹰的作品定位为欧美对非白人美学的评论。例如,她的油画《无题(珊瑚色、绿松石色和黄色)》(2016),在橙色和红色的颜料带中嵌入了一排排复杂的蓝色和黄色珠子,让人联想到色彩场画。白鹰在博物馆的主要展览空间中占据了最重要的位置,四个房间展示了艺术家在多种媒体上的技巧:丙烯画、珠饰、视频和摄影。她在那些与欧美艺术家有关的作品中层层叠加并利用土著主题。这次展览探讨了关于艺术与工艺的错误二元对立以及土著妇女在美洲土著社区中的中心地位的争论;这次展览巧妙地处理了前者,而后者似乎没有得到充分的探索。
{"title":"Dyani White Hawk: Speaking to Relatives, Kemper Museum of Contemporary Art, Kansas City, MO, 18 February–16 May 2021","authors":"Glynnis Stevenson","doi":"10.14324/111.444.2396-9008.1340","DOIUrl":"https://doi.org/10.14324/111.444.2396-9008.1340","url":null,"abstract":"This review looks at the Kemper Museum of Contemporary Art’s tribute to the past decade of artwork by Dyani White Hawk, a curator and multimedia artist of Sičáŋǧu Lakota, German and Welsh ancestry. The exhibition entrance sits directly across from Sean Scully’s The Moroccan (1984), positioning White Hawk’s work as a commentary on the Euro-American appropriation of non-white aesthetics. See for example, her canvas Untitled (coral, turquoise, and yellow) (2016), which embeds intricate rows of blue and yellow beads within bands of orange and red paint that evoke colour field painting. White Hawk receives pride of place in the museum’s main exhibition space, a series of four rooms showcasing the artist’s skill in many media: acrylic painting, beadwork, video, and photography. She layers and leverages Indigenous motifs amidst those associated with Euro-American artists. The exhibition tackles the debate over the false binary of art vs. craft and the centrality of indigenous women to Native American communities; the exhibition tackles the former deftly, the latter feels less fully explored.","PeriodicalId":72958,"journal":{"name":"Eurographics ... Workshop on 3D Object Retrieval : EG 3DOR. Eurographics Workshop on 3D Object Retrieval","volume":"41 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77941287","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 : 2022-01-27DOI: 10.14324/111.444.2396-9008.1337
Kitty Whittell
This article considers the role of fluidity in two early liquid-based works by Hans Haacke and Gustav Metzger, Condensation Cube (1965) and Liquid Crystal Environment (1965). Fluidity is distinct from liquidity in that it presents a process of transition from solid to liquid in the case of Liquid Crystal Environment and liquid to gas in Condensation Cube. The stakes of this fluid process are, as I argue, crucial to understanding the critical stakes of these works in relation to Haacke and Metzger's artistic practise. By examining fluid process, I hope to show how these early pieces also contribute institutional and political criticisms that are typically associated with Haacke and Metzger's later work.
{"title":"Investigating Fluidity in Hans Haacke’s Condensation Cube (1965) and Gustave Metzger’s Liquid Crystal Environment (1965)","authors":"Kitty Whittell","doi":"10.14324/111.444.2396-9008.1337","DOIUrl":"https://doi.org/10.14324/111.444.2396-9008.1337","url":null,"abstract":"This article considers the role of fluidity in two early liquid-based works by Hans Haacke and Gustav Metzger, Condensation Cube (1965) and Liquid Crystal Environment (1965). Fluidity is distinct from liquidity in that it presents a process of transition from solid to liquid in the case of Liquid Crystal Environment and liquid to gas in Condensation Cube. The stakes of this fluid process are, as I argue, crucial to understanding the critical stakes of these works in relation to Haacke and Metzger's artistic practise. By examining fluid process, I hope to show how these early pieces also contribute institutional and political criticisms that are typically associated with Haacke and Metzger's later work.","PeriodicalId":72958,"journal":{"name":"Eurographics ... Workshop on 3D Object Retrieval : EG 3DOR. Eurographics Workshop on 3D Object Retrieval","volume":"5 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85712116","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 : 2022-01-27DOI: 10.14324/111.444.2396-9008.1341
Louis Shankar
In 2020, Tate Modern hosted a retrospective of Andy Warhol curated by Gregor Muir and Yilmaz Dziewior. They aimed to focus on the life of the artist, looking to his family, his sexuality, and his beliefs. This review assesses these aims, focusing on the queer strategies at work in recent Warhol scholarship.
{"title":"Andy Warhol, Tate Modern, London, 12 March–15 November 2020","authors":"Louis Shankar","doi":"10.14324/111.444.2396-9008.1341","DOIUrl":"https://doi.org/10.14324/111.444.2396-9008.1341","url":null,"abstract":"In 2020, Tate Modern hosted a retrospective of Andy Warhol curated by Gregor Muir and Yilmaz Dziewior. They aimed to focus on the life of the artist, looking to his family, his sexuality, and his beliefs. This review assesses these aims, focusing on the queer strategies at work in recent Warhol scholarship.","PeriodicalId":72958,"journal":{"name":"Eurographics ... Workshop on 3D Object Retrieval : EG 3DOR. Eurographics Workshop on 3D Object Retrieval","volume":"2013 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83160185","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 : 2022-01-27DOI: 10.14324/111.444.2396-9008.1339
T. Sheikhan
During her lifetime, Juliette Récamier (1777-1849) was widely considered as the most beautiful woman in France. Within the country’s post-revolutionary Directory and Consulate epoch, Récamier ran a salon that boasted prominent political figures and writers among its frequent guests. Récamier’s own legacy however is one of seemingly total passivity: praise aside, the impression we are left with is one of placid docility. Récamier was greatly admired, yet incapable of articulating her own thoughts and ideas. This article examines François Gérard’s portrait of Récamier. At the time, Gérard (1770-1837) was one of the period’s most successful artists, whose services Napoléon would request for his coronation portrait in 1805. Gérard’s work leads to this study’s argument that the articulation and presentation of self image in the early nineteenth century was no less layered and complex than the creation of material images. In fact, it illustrates just how much Gérard’s portrait is an image of her making as much as it is of his. We see such agency further mediated in Récamier’s shawl dance, a performative staple of the soirées she hosted. Imbuing the artistic virtues associated with neoclassical dress with the passion, sensuality, and individualistic nature of dance; this article considers Récamier’s self-fashioning on a stage exclusively within her salon.
{"title":"Politics, Fashion and Female Agency in Parisian Salons c. 1800: The Case of Juliette Récamier","authors":"T. Sheikhan","doi":"10.14324/111.444.2396-9008.1339","DOIUrl":"https://doi.org/10.14324/111.444.2396-9008.1339","url":null,"abstract":"During her lifetime, Juliette Récamier (1777-1849) was widely considered as the most beautiful woman in France. Within the country’s post-revolutionary Directory and Consulate epoch, Récamier ran a salon that boasted prominent political figures and writers among its frequent guests. Récamier’s own legacy however is one of seemingly total passivity: praise aside, the impression we are left with is one of placid docility. Récamier was greatly admired, yet incapable of articulating her own thoughts and ideas. This article examines François Gérard’s portrait of Récamier. At the time, Gérard (1770-1837) was one of the period’s most successful artists, whose services Napoléon would request for his coronation portrait in 1805. Gérard’s work leads to this study’s argument that the articulation and presentation of self image in the early nineteenth century was no less layered and complex than the creation of material images. In fact, it illustrates just how much Gérard’s portrait is an image of her making as much as it is of his. We see such agency further mediated in Récamier’s shawl dance, a performative staple of the soirées she hosted. Imbuing the artistic virtues associated with neoclassical dress with the passion, sensuality, and individualistic nature of dance; this article considers Récamier’s self-fashioning on a stage exclusively within her salon.","PeriodicalId":72958,"journal":{"name":"Eurographics ... Workshop on 3D Object Retrieval : EG 3DOR. Eurographics Workshop on 3D Object Retrieval","volume":"40 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86336570","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 : 2022-01-26DOI: 10.14324/111.444.2396-9008.1303
Not Applicable
Theses in Progress
正在进行的论文
{"title":"Theses in Progress","authors":"Not Applicable","doi":"10.14324/111.444.2396-9008.1303","DOIUrl":"https://doi.org/10.14324/111.444.2396-9008.1303","url":null,"abstract":"Theses in Progress","PeriodicalId":72958,"journal":{"name":"Eurographics ... Workshop on 3D Object Retrieval : EG 3DOR. Eurographics Workshop on 3D Object Retrieval","volume":"22 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76557305","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}
Meitar Shechter, R. Hanocka, G. Metzer, R. Giryes, D. Cohen-Or
We introduce NeuralMLS, a space-based deformation technique, guided by a set of displaced control points. We leverage the power of neural networks to inject the underlying shape geometry into the deformation parameters. The goal of our technique is to enable a realistic and intuitive shape deformation. Our method is built upon moving least-squares (MLS), since it minimizes a weighted sum of the given control point displacements. Traditionally, the influence of each control point on every point in space (i.e., the weighting function) is defined using inverse distance heuristics. In this work, we opt to learn the weighting function, by training a neural network on the control points from a single input shape, and exploit the innate smoothness of neural networks. Our geometry-aware control point deformation is agnostic to the surface representation and quality; it can be applied to point clouds or meshes, including non-manifold and disconnected surface soups. We show that our technique facilitates intuitive piecewise smooth deformations, which are well suited for manufactured objects. We show the advantages of our approach compared to existing surface and space-based deformation techniques, both quantitatively and qualitatively.
{"title":"NeuralMLS: Geometry-Aware Control Point Deformation","authors":"Meitar Shechter, R. Hanocka, G. Metzer, R. Giryes, D. Cohen-Or","doi":"10.2312/egs.20221034","DOIUrl":"https://doi.org/10.2312/egs.20221034","url":null,"abstract":"We introduce NeuralMLS, a space-based deformation technique, guided by a set of displaced control points. We leverage the power of neural networks to inject the underlying shape geometry into the deformation parameters. The goal of our technique is to enable a realistic and intuitive shape deformation. Our method is built upon moving least-squares (MLS), since it minimizes a weighted sum of the given control point displacements. Traditionally, the influence of each control point on every point in space (i.e., the weighting function) is defined using inverse distance heuristics. In this work, we opt to learn the weighting function, by training a neural network on the control points from a single input shape, and exploit the innate smoothness of neural networks. Our geometry-aware control point deformation is agnostic to the surface representation and quality; it can be applied to point clouds or meshes, including non-manifold and disconnected surface soups. We show that our technique facilitates intuitive piecewise smooth deformations, which are well suited for manufactured objects. We show the advantages of our approach compared to existing surface and space-based deformation techniques, both quantitatively and qualitatively.","PeriodicalId":72958,"journal":{"name":"Eurographics ... Workshop on 3D Object Retrieval : EG 3DOR. Eurographics Workshop on 3D Object Retrieval","volume":"31 1","pages":"65-68"},"PeriodicalIF":0.0,"publicationDate":"2022-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78405793","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}
We propose an efficient conservative cone map generation algorithm that has Θ ( N 2 log N ) complexity for textures of dimension N × N in contrast to the Θ ( N 4 ) complexity of brute-force approaches. This is achieved by using a maximum mip texture of a heightmap to process all texels during the search for cone apertures, resulting in real-time generation times. Furthermore, we show that discarding already visited regions of neighboring mip texels widens the obtained cones considerably while still being conservative. Finally, we present a method to increase cone aperture tangents further at the expense of conservativeness. We compare our methods to brute-force and relaxed cone maps in generation and rendering performance.
{"title":"Quick Cone Map Generation on the GPU","authors":"Gábor Valasek, Róbert Bán","doi":"10.2312/egs.20221021","DOIUrl":"https://doi.org/10.2312/egs.20221021","url":null,"abstract":"We propose an efficient conservative cone map generation algorithm that has Θ ( N 2 log N ) complexity for textures of dimension N × N in contrast to the Θ ( N 4 ) complexity of brute-force approaches. This is achieved by using a maximum mip texture of a heightmap to process all texels during the search for cone apertures, resulting in real-time generation times. Furthermore, we show that discarding already visited regions of neighboring mip texels widens the obtained cones considerably while still being conservative. Finally, we present a method to increase cone aperture tangents further at the expense of conservativeness. We compare our methods to brute-force and relaxed cone maps in generation and rendering performance.","PeriodicalId":72958,"journal":{"name":"Eurographics ... Workshop on 3D Object Retrieval : EG 3DOR. Eurographics Workshop on 3D Object Retrieval","volume":"20 1","pages":"13-16"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77745288","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}
Tristan Wirth, Aria Jamili, Max von Bülow, V. Knauthe, S. Guthe
Due to material properties, monocular depth estimation of transparent structures is inherently challenging. Recent advances leverage additional knowledge that is not available in all contexts, i.e., known shape or depth information from a sensor. General-purpose machine learning models, that do not utilize such additional knowledge, have not yet been explicitly evaluated regarding their performance on transparent structures. In this work, we show that these models show poor performance on the depth estimation of transparent structures. However, fine-tuning on suitable data sets, such as ClearGrasp, increases their estimation performance on the task at hand. Our evaluations show that high performance on general-purpose benchmarks translates well into performance on transparent objects after fine-tuning. Furthermore, our analysis suggests, that state-of-theart high-performing models are not able to capture a high grade of detail from both the image foreground and background at the same time. This finding shows the demand for a combination of existing models to further enhance depth estimation quality. CCS Concepts • Computing methodologies → Computer vision; Shape inference;
{"title":"Fitness of General-Purpose Monocular Depth Estimation Architectures for Transparent Structures","authors":"Tristan Wirth, Aria Jamili, Max von Bülow, V. Knauthe, S. Guthe","doi":"10.2312/egs.20221020","DOIUrl":"https://doi.org/10.2312/egs.20221020","url":null,"abstract":"Due to material properties, monocular depth estimation of transparent structures is inherently challenging. Recent advances leverage additional knowledge that is not available in all contexts, i.e., known shape or depth information from a sensor. General-purpose machine learning models, that do not utilize such additional knowledge, have not yet been explicitly evaluated regarding their performance on transparent structures. In this work, we show that these models show poor performance on the depth estimation of transparent structures. However, fine-tuning on suitable data sets, such as ClearGrasp, increases their estimation performance on the task at hand. Our evaluations show that high performance on general-purpose benchmarks translates well into performance on transparent objects after fine-tuning. Furthermore, our analysis suggests, that state-of-theart high-performing models are not able to capture a high grade of detail from both the image foreground and background at the same time. This finding shows the demand for a combination of existing models to further enhance depth estimation quality. CCS Concepts • Computing methodologies → Computer vision; Shape inference;","PeriodicalId":72958,"journal":{"name":"Eurographics ... Workshop on 3D Object Retrieval : EG 3DOR. Eurographics Workshop on 3D Object Retrieval","volume":"17 1","pages":"9-12"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81302779","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}