Zhuheng Lu, Ting Wu, Yuewei Dai, Weiqing Li, Zhiyong Su
Two forms of imbalances are commonly observed in point cloud semantic segmentation datasets: (1) category imbalances, where certain objects are more prevalent than others; and (2) size imbalances, where certain objects occupy more points than others. Because of this, the majority of categories and large objects are favored in the existing evaluation metrics. This paper suggests fine-grained mIoU and mAcc for a more thorough assessment of point cloud segmentation algorithms in order to address these issues. Richer statistical information is provided for models and datasets by these fine-grained metrics, which also lessen the bias of current semantic segmentation metrics towards large objects. The proposed metrics are used to train and assess various semantic segmentation algorithms on three distinct indoor and outdoor semantic segmentation datasets.
{"title":"Fine-grained Metrics for Point Cloud Semantic Segmentation","authors":"Zhuheng Lu, Ting Wu, Yuewei Dai, Weiqing Li, Zhiyong Su","doi":"arxiv-2407.21289","DOIUrl":"https://doi.org/arxiv-2407.21289","url":null,"abstract":"Two forms of imbalances are commonly observed in point cloud semantic\u0000segmentation datasets: (1) category imbalances, where certain objects are more\u0000prevalent than others; and (2) size imbalances, where certain objects occupy\u0000more points than others. Because of this, the majority of categories and large\u0000objects are favored in the existing evaluation metrics. This paper suggests\u0000fine-grained mIoU and mAcc for a more thorough assessment of point cloud\u0000segmentation algorithms in order to address these issues. Richer statistical\u0000information is provided for models and datasets by these fine-grained metrics,\u0000which also lessen the bias of current semantic segmentation metrics towards\u0000large objects. The proposed metrics are used to train and assess various\u0000semantic segmentation algorithms on three distinct indoor and outdoor semantic\u0000segmentation datasets.","PeriodicalId":501174,"journal":{"name":"arXiv - CS - Graphics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863840","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}
This comparative study evaluates various neural surface reconstruction methods, particularly focusing on their implications for scientific visualization through reconstructing 3D surfaces via multi-view rendering images. We categorize ten methods into neural radiance fields and neural implicit surfaces, uncovering the benefits of leveraging distance functions (i.e., SDFs and UDFs) to enhance the accuracy and smoothness of the reconstructed surfaces. Our findings highlight the efficiency and quality of NeuS2 for reconstructing closed surfaces and identify NeUDF as a promising candidate for reconstructing open surfaces despite some limitations. By sharing our benchmark dataset, we invite researchers to test the performance of their methods, contributing to the advancement of surface reconstruction solutions for scientific visualization.
{"title":"A Comparative Study of Neural Surface Reconstruction for Scientific Visualization","authors":"Siyuan Yao, Weixi Song, Chaoli Wang","doi":"arxiv-2407.20868","DOIUrl":"https://doi.org/arxiv-2407.20868","url":null,"abstract":"This comparative study evaluates various neural surface reconstruction\u0000methods, particularly focusing on their implications for scientific\u0000visualization through reconstructing 3D surfaces via multi-view rendering\u0000images. We categorize ten methods into neural radiance fields and neural\u0000implicit surfaces, uncovering the benefits of leveraging distance functions\u0000(i.e., SDFs and UDFs) to enhance the accuracy and smoothness of the\u0000reconstructed surfaces. Our findings highlight the efficiency and quality of\u0000NeuS2 for reconstructing closed surfaces and identify NeUDF as a promising\u0000candidate for reconstructing open surfaces despite some limitations. By sharing\u0000our benchmark dataset, we invite researchers to test the performance of their\u0000methods, contributing to the advancement of surface reconstruction solutions\u0000for scientific visualization.","PeriodicalId":501174,"journal":{"name":"arXiv - CS - Graphics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863844","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}
Learning the prior knowledge of the 3D human-object spatial relation is crucial for reconstructing human-object interaction from images and understanding how humans interact with objects in 3D space. Previous works learn this prior from datasets collected in controlled environments, but due to the diversity of domains, they struggle to generalize to real-world scenarios. To overcome this limitation, we present a 2D-supervised method that learns the 3D human-object spatial relation prior purely from 2D images in the wild. Our method utilizes a flow-based neural network to learn the prior distribution of the 2D human-object keypoint layout and viewports for each image in the dataset. The effectiveness of the prior learned from 2D images is demonstrated on the human-object reconstruction task by applying the prior to tune the relative pose between the human and the object during the post-optimization stage. To validate and benchmark our method on in-the-wild images, we collect the WildHOI dataset from the YouTube website, which consists of various interactions with 8 objects in real-world scenarios. We conduct the experiments on the indoor BEHAVE dataset and the outdoor WildHOI dataset. The results show that our method achieves almost comparable performance with fully 3D supervised methods on the BEHAVE dataset, even if we have only utilized the 2D layout information, and outperforms previous methods in terms of generality and interaction diversity on in-the-wild images.
{"title":"Monocular Human-Object Reconstruction in the Wild","authors":"Chaofan Huo, Ye Shi, Jingya Wang","doi":"arxiv-2407.20566","DOIUrl":"https://doi.org/arxiv-2407.20566","url":null,"abstract":"Learning the prior knowledge of the 3D human-object spatial relation is\u0000crucial for reconstructing human-object interaction from images and\u0000understanding how humans interact with objects in 3D space. Previous works\u0000learn this prior from datasets collected in controlled environments, but due to\u0000the diversity of domains, they struggle to generalize to real-world scenarios.\u0000To overcome this limitation, we present a 2D-supervised method that learns the\u00003D human-object spatial relation prior purely from 2D images in the wild. Our\u0000method utilizes a flow-based neural network to learn the prior distribution of\u0000the 2D human-object keypoint layout and viewports for each image in the\u0000dataset. The effectiveness of the prior learned from 2D images is demonstrated\u0000on the human-object reconstruction task by applying the prior to tune the\u0000relative pose between the human and the object during the post-optimization\u0000stage. To validate and benchmark our method on in-the-wild images, we collect\u0000the WildHOI dataset from the YouTube website, which consists of various\u0000interactions with 8 objects in real-world scenarios. We conduct the experiments\u0000on the indoor BEHAVE dataset and the outdoor WildHOI dataset. The results show\u0000that our method achieves almost comparable performance with fully 3D supervised\u0000methods on the BEHAVE dataset, even if we have only utilized the 2D layout\u0000information, and outperforms previous methods in terms of generality and\u0000interaction diversity on in-the-wild images.","PeriodicalId":501174,"journal":{"name":"arXiv - CS - Graphics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863845","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}
Chaofan Huo, Ye Shi, Yuexin Ma, Lan Xu, Jingyi Yu, Jingya Wang
Modeling and capturing the 3D spatial arrangement of the human and the object is the key to perceiving 3D human-object interaction from monocular images. In this work, we propose to use the Human-Object Offset between anchors which are densely sampled from the surface of human mesh and object mesh to represent human-object spatial relation. Compared with previous works which use contact map or implicit distance filed to encode 3D human-object spatial relations, our method is a simple and efficient way to encode the highly detailed spatial correlation between the human and object. Based on this representation, we propose Stacked Normalizing Flow (StackFLOW) to infer the posterior distribution of human-object spatial relations from the image. During the optimization stage, we finetune the human body pose and object 6D pose by maximizing the likelihood of samples based on this posterior distribution and minimizing the 2D-3D corresponding reprojection loss. Extensive experimental results show that our method achieves impressive results on two challenging benchmarks, BEHAVE and InterCap datasets.
{"title":"StackFLOW: Monocular Human-Object Reconstruction by Stacked Normalizing Flow with Offset","authors":"Chaofan Huo, Ye Shi, Yuexin Ma, Lan Xu, Jingyi Yu, Jingya Wang","doi":"arxiv-2407.20545","DOIUrl":"https://doi.org/arxiv-2407.20545","url":null,"abstract":"Modeling and capturing the 3D spatial arrangement of the human and the object\u0000is the key to perceiving 3D human-object interaction from monocular images. In\u0000this work, we propose to use the Human-Object Offset between anchors which are\u0000densely sampled from the surface of human mesh and object mesh to represent\u0000human-object spatial relation. Compared with previous works which use contact\u0000map or implicit distance filed to encode 3D human-object spatial relations, our\u0000method is a simple and efficient way to encode the highly detailed spatial\u0000correlation between the human and object. Based on this representation, we\u0000propose Stacked Normalizing Flow (StackFLOW) to infer the posterior\u0000distribution of human-object spatial relations from the image. During the\u0000optimization stage, we finetune the human body pose and object 6D pose by\u0000maximizing the likelihood of samples based on this posterior distribution and\u0000minimizing the 2D-3D corresponding reprojection loss. Extensive experimental\u0000results show that our method achieves impressive results on two challenging\u0000benchmarks, BEHAVE and InterCap datasets.","PeriodicalId":501174,"journal":{"name":"arXiv - CS - Graphics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863847","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}
The Python colorspace package provides a toolbox for mapping between different color spaces which can then be used to generate a wide range of perceptually-based color palettes for qualitative or quantitative (sequential or diverging) information. These palettes (as well as any other sets of colors) can be visualized, assessed, and manipulated in various ways, e.g., by color swatches, emulating the effects of color vision deficiencies, or depicting the perceptual properties. Finally, the color palettes generated by the package can be easily integrated into standard visualization workflows in Python, e.g., using matplotlib, seaborn, or plotly.
{"title":"colorspace: A Python Toolbox for Manipulating and Assessing Colors and Palettes","authors":"Reto Stauffer, Achim Zeileis","doi":"arxiv-2407.19921","DOIUrl":"https://doi.org/arxiv-2407.19921","url":null,"abstract":"The Python colorspace package provides a toolbox for mapping between\u0000different color spaces which can then be used to generate a wide range of\u0000perceptually-based color palettes for qualitative or quantitative (sequential\u0000or diverging) information. These palettes (as well as any other sets of colors)\u0000can be visualized, assessed, and manipulated in various ways, e.g., by color\u0000swatches, emulating the effects of color vision deficiencies, or depicting the\u0000perceptual properties. Finally, the color palettes generated by the package can\u0000be easily integrated into standard visualization workflows in Python, e.g.,\u0000using matplotlib, seaborn, or plotly.","PeriodicalId":501174,"journal":{"name":"arXiv - CS - Graphics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863850","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}
Hypergraphs provide a natural way to represent polyadic relationships in network data. For large hypergraphs, it is often difficult to visually detect structures within the data. Recently, a scalable polygon-based visualization approach was developed allowing hypergraphs with thousands of hyperedges to be simplified and examined at different levels of detail. However, this approach is not guaranteed to eliminate all of the visual clutter caused by unavoidable overlaps. Furthermore, meaningful structures can be lost at simplified scales, making their interpretation unreliable. In this paper, we define hypergraph structures using the bipartite graph representation, allowing us to decompose the hypergraph into a union of structures including topological blocks, bridges, and branches, and to identify exactly where unavoidable overlaps must occur. We also introduce a set of topology preserving and topology altering atomic operations, enabling the preservation of important structures while reducing unavoidable overlaps to improve visual clarity and interpretability in simplified scales. We demonstrate our approach in several real-world applications.
{"title":"Structure-Aware Simplification for Hypergraph Visualization","authors":"Peter Oliver, Eugene Zhang, Yue Zhang","doi":"arxiv-2407.19621","DOIUrl":"https://doi.org/arxiv-2407.19621","url":null,"abstract":"Hypergraphs provide a natural way to represent polyadic relationships in\u0000network data. For large hypergraphs, it is often difficult to visually detect\u0000structures within the data. Recently, a scalable polygon-based visualization\u0000approach was developed allowing hypergraphs with thousands of hyperedges to be\u0000simplified and examined at different levels of detail. However, this approach\u0000is not guaranteed to eliminate all of the visual clutter caused by unavoidable\u0000overlaps. Furthermore, meaningful structures can be lost at simplified scales,\u0000making their interpretation unreliable. In this paper, we define hypergraph\u0000structures using the bipartite graph representation, allowing us to decompose\u0000the hypergraph into a union of structures including topological blocks,\u0000bridges, and branches, and to identify exactly where unavoidable overlaps must\u0000occur. We also introduce a set of topology preserving and topology altering\u0000atomic operations, enabling the preservation of important structures while\u0000reducing unavoidable overlaps to improve visual clarity and interpretability in\u0000simplified scales. We demonstrate our approach in several real-world\u0000applications.","PeriodicalId":501174,"journal":{"name":"arXiv - CS - Graphics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863851","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}
In the field of architecture, the conversion of single images into 2 and 1/2D and 3D meshes is a promising technology that enhances design visualization and efficiency. This paper evaluates four innovative methods: "One-2-3-45," "CRM: Single Image to 3D Textured Mesh with Convolutional Reconstruction Model," "Instant Mesh," and "Image-to-Mesh." These methods are at the forefront of this technology, focusing on their applicability in architectural design and visualization. They streamline the creation of 3D architectural models, enabling rapid prototyping and detailed visualization from minimal initial inputs, such as photographs or simple sketches.One-2-3-45 leverages a diffusion-based approach to generate multi-view reconstructions, ensuring high geometric fidelity and texture quality. CRM utilizes a convolutional network to integrate geometric priors into its architecture, producing detailed and textured meshes quickly and efficiently. Instant Mesh combines the strengths of multi-view diffusion and sparse-view models to offer speed and scalability, suitable for diverse architectural projects. Image-to-Mesh leverages a generative adversarial network (GAN) to produce 3D meshes from single images, focusing on maintaining high texture fidelity and geometric accuracy by incorporating image and depth map data into its training process. It uses a hybrid approach that combines voxel-based representations with surface reconstruction techniques to ensure detailed and realistic 3D models.This comparative study highlights each method's contribution to reducing design cycle times, improving accuracy, and enabling flexible adaptations to various architectural styles and requirements. By providing architects with powerful tools for rapid visualization and iteration, these advancements in 3D mesh generation are set to revolutionize architectural practices.
{"title":"From Flat to Spatial: Comparison of 4 methods constructing 3D, 2 and 1/2D Models from 2D Plans with neural networks","authors":"Jacob Sam, Karan Patel, Mike Saad","doi":"arxiv-2407.19970","DOIUrl":"https://doi.org/arxiv-2407.19970","url":null,"abstract":"In the field of architecture, the conversion of single images into 2 and 1/2D\u0000and 3D meshes is a promising technology that enhances design visualization and\u0000efficiency. This paper evaluates four innovative methods: \"One-2-3-45,\" \"CRM:\u0000Single Image to 3D Textured Mesh with Convolutional Reconstruction Model,\"\u0000\"Instant Mesh,\" and \"Image-to-Mesh.\" These methods are at the forefront of this\u0000technology, focusing on their applicability in architectural design and\u0000visualization. They streamline the creation of 3D architectural models,\u0000enabling rapid prototyping and detailed visualization from minimal initial\u0000inputs, such as photographs or simple sketches.One-2-3-45 leverages a\u0000diffusion-based approach to generate multi-view reconstructions, ensuring high\u0000geometric fidelity and texture quality. CRM utilizes a convolutional network to\u0000integrate geometric priors into its architecture, producing detailed and\u0000textured meshes quickly and efficiently. Instant Mesh combines the strengths of\u0000multi-view diffusion and sparse-view models to offer speed and scalability,\u0000suitable for diverse architectural projects. Image-to-Mesh leverages a\u0000generative adversarial network (GAN) to produce 3D meshes from single images,\u0000focusing on maintaining high texture fidelity and geometric accuracy by\u0000incorporating image and depth map data into its training process. It uses a\u0000hybrid approach that combines voxel-based representations with surface\u0000reconstruction techniques to ensure detailed and realistic 3D models.This\u0000comparative study highlights each method's contribution to reducing design\u0000cycle times, improving accuracy, and enabling flexible adaptations to various\u0000architectural styles and requirements. By providing architects with powerful\u0000tools for rapid visualization and iteration, these advancements in 3D mesh\u0000generation are set to revolutionize architectural practices.","PeriodicalId":501174,"journal":{"name":"arXiv - CS - Graphics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863849","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}
This work presents a web-based, open-source path tracer for rendering physically-based 3D scenes using WebGPU and the OpenPBR surface shading model. While rasterization has been the dominant real-time rendering technique on the web since WebGL's introduction in 2011, it struggles with global illumination. This necessitates more complex techniques, often relying on pregenerated artifacts to attain the desired level of visual fidelity. Path tracing inherently addresses these limitations but at the cost of increased rendering time. Our work focuses on industrial applications where highly customizable products are common and real-time performance is not critical. We leverage WebGPU to implement path tracing on the web, integrating the OpenPBR standard for physically-based material representation. The result is a near real-time path tracer capable of rendering high-fidelity 3D scenes directly in web browsers, eliminating the need for pregenerated assets. Our implementation demonstrates the potential of WebGPU for advanced rendering techniques and opens new possibilities for web-based 3D visualization in industrial applications.
{"title":"Physically-based Path Tracer using WebGPU and OpenPBR","authors":"Simon Stucki, Philipp Ackermann","doi":"arxiv-2407.19977","DOIUrl":"https://doi.org/arxiv-2407.19977","url":null,"abstract":"This work presents a web-based, open-source path tracer for rendering\u0000physically-based 3D scenes using WebGPU and the OpenPBR surface shading model.\u0000While rasterization has been the dominant real-time rendering technique on the\u0000web since WebGL's introduction in 2011, it struggles with global illumination.\u0000This necessitates more complex techniques, often relying on pregenerated\u0000artifacts to attain the desired level of visual fidelity. Path tracing\u0000inherently addresses these limitations but at the cost of increased rendering\u0000time. Our work focuses on industrial applications where highly customizable\u0000products are common and real-time performance is not critical. We leverage\u0000WebGPU to implement path tracing on the web, integrating the OpenPBR standard\u0000for physically-based material representation. The result is a near real-time\u0000path tracer capable of rendering high-fidelity 3D scenes directly in web\u0000browsers, eliminating the need for pregenerated assets. Our implementation\u0000demonstrates the potential of WebGPU for advanced rendering techniques and\u0000opens new possibilities for web-based 3D visualization in industrial\u0000applications.","PeriodicalId":501174,"journal":{"name":"arXiv - CS - Graphics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863848","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}
Chengan He, Xin Sun, Zhixin Shu, Fujun Luan, Sören Pirk, Jorge Alejandro Amador Herrera, Dominik L. Michels, Tuanfeng Y. Wang, Meng Zhang, Holly Rushmeier, Yi Zhou
We present textsc{Perm}, a learned parametric model of human 3D hair designed to facilitate various hair-related applications. Unlike previous work that jointly models the global hair shape and local strand details, we propose to disentangle them using a PCA-based strand representation in the frequency domain, thereby allowing more precise editing and output control. Specifically, we leverage our strand representation to fit and decompose hair geometry textures into low- to high-frequency hair structures. These decomposed textures are later parameterized with different generative models, emulating common stages in the hair modeling process. We conduct extensive experiments to validate the architecture design of textsc{Perm}, and finally deploy the trained model as a generic prior to solve task-agnostic problems, further showcasing its flexibility and superiority in tasks such as 3D hair parameterization, hairstyle interpolation, single-view hair reconstruction, and hair-conditioned image generation. Our code and data will be available at: url{https://github.com/c-he/perm}.
{"title":"textsc{Perm}: A Parametric Representation for Multi-Style 3D Hair Modeling","authors":"Chengan He, Xin Sun, Zhixin Shu, Fujun Luan, Sören Pirk, Jorge Alejandro Amador Herrera, Dominik L. Michels, Tuanfeng Y. Wang, Meng Zhang, Holly Rushmeier, Yi Zhou","doi":"arxiv-2407.19451","DOIUrl":"https://doi.org/arxiv-2407.19451","url":null,"abstract":"We present textsc{Perm}, a learned parametric model of human 3D hair\u0000designed to facilitate various hair-related applications. Unlike previous work\u0000that jointly models the global hair shape and local strand details, we propose\u0000to disentangle them using a PCA-based strand representation in the frequency\u0000domain, thereby allowing more precise editing and output control. Specifically,\u0000we leverage our strand representation to fit and decompose hair geometry\u0000textures into low- to high-frequency hair structures. These decomposed textures\u0000are later parameterized with different generative models, emulating common\u0000stages in the hair modeling process. We conduct extensive experiments to\u0000validate the architecture design of textsc{Perm}, and finally deploy the\u0000trained model as a generic prior to solve task-agnostic problems, further\u0000showcasing its flexibility and superiority in tasks such as 3D hair\u0000parameterization, hairstyle interpolation, single-view hair reconstruction, and\u0000hair-conditioned image generation. Our code and data will be available at:\u0000url{https://github.com/c-he/perm}.","PeriodicalId":501174,"journal":{"name":"arXiv - CS - Graphics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863984","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}
Freeform thin-shell surfaces are critical in various fields, but their fabrication is complex and costly. Traditional methods are wasteful and require custom molds, while 3D printing needs extensive support structures and post-processing. Thermoshrinkage actuated 4D printing is an effective method through flat structures fabricating 3D shell. However, existing research faces issues related to precise deformation and limited robustness. Addressing these issues is challenging due to three key factors: (1) Difficulty in finding a universal method to control deformation across different materials; (2) Variability in deformation influenced by factors such as printing speed, layer thickness, and heating temperature; (3) Environmental factors affecting the deformation process. To overcome these challenges, we introduce FreeShell, a robust 4D printing technique that uses thermoshrinkage to create precise 3D shells. This method prints triangular tiles connected by shrinkable connectors from a single material. Upon heating, the connectors shrink, moving the tiles to form the desired 3D shape, simplifying fabrication and reducing material and environment dependency. An optimized algorithm for flattening 3D meshes ensures precision in printing. FreeShell demonstrates its effectiveness through various examples and experiments, showcasing accuracy, robustness, and strength, representing advancement in fabricating complex freeform surfaces.
{"title":"FreeShell: A Context-Free 4D Printing Technique for Fabricating Complex 3D Triangle Mesh Shells","authors":"Chao Yuan, Nan Cao, Xuejiao Ma, Shengqi Dang","doi":"arxiv-2407.19533","DOIUrl":"https://doi.org/arxiv-2407.19533","url":null,"abstract":"Freeform thin-shell surfaces are critical in various fields, but their\u0000fabrication is complex and costly. Traditional methods are wasteful and require\u0000custom molds, while 3D printing needs extensive support structures and\u0000post-processing. Thermoshrinkage actuated 4D printing is an effective method\u0000through flat structures fabricating 3D shell. However, existing research faces\u0000issues related to precise deformation and limited robustness. Addressing these\u0000issues is challenging due to three key factors: (1) Difficulty in finding a\u0000universal method to control deformation across different materials; (2)\u0000Variability in deformation influenced by factors such as printing speed, layer\u0000thickness, and heating temperature; (3) Environmental factors affecting the\u0000deformation process. To overcome these challenges, we introduce FreeShell, a\u0000robust 4D printing technique that uses thermoshrinkage to create precise 3D\u0000shells. This method prints triangular tiles connected by shrinkable connectors\u0000from a single material. Upon heating, the connectors shrink, moving the tiles\u0000to form the desired 3D shape, simplifying fabrication and reducing material and\u0000environment dependency. An optimized algorithm for flattening 3D meshes ensures\u0000precision in printing. FreeShell demonstrates its effectiveness through various\u0000examples and experiments, showcasing accuracy, robustness, and strength,\u0000representing advancement in fabricating complex freeform surfaces.","PeriodicalId":501174,"journal":{"name":"arXiv - CS - Graphics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141863853","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：