【深度学习】深度学习三维人体建模最新论文、资源、数据、代码整理分享

    三维人体建模作为计算机人体仿真的一个组成部分，一直是人们研究的热点之一。自交互式计算机图形学诞生之日起，就有学者不断探索计算机人体建模技术。从线框建模、实体建模、曲面建模发展到基于物理的建模，已取得重大进展。3维度人体建模在医学图像、生物医学、手势识别、视频会议、视频游戏、自动新闻播放、电影制作、材料变形、图象压缩等方面都有实际应用价值。

    本资源整理了最近几年基于深度学习技术的三维人体重构（建模）相关的最新论文、资源、代码和公开数据集，分享给需要的朋友。

    资源整理自网络，源地址：https://github.com/chenweikai/Body_Reconstruction_References

目录

论文资源列表

Multi-view Reconstruction

    1. Deep Learning

    [ICCV19] Shape-Aware Human Pose and Shape Reconstruction Using Multi-View Images 

    [ICCV19] TexturePose: Supervising Human Mesh Estimation with Texture Consistency 

    [ICCV19] Human Mesh Recovery From Monocular Images via a Skeleton-Disentangled Representation

    [ECCV18] Deep Volumetric Video From Very Sparse Multi-View Performance Capture 

    [ECCV18] Bodynet: Volumetric inference of 3d human body shapes 

    [ECCV18] Volumetric performance capture from minimal camera viewpoints 

    2. Shape from Silhouette

    [CGF16] 3d body shapes estimation from dressed-human silhouettes 

    [ICCV15] Interactive visual hull refinement for specular and transparent object surface reconstruction

    [HPG13] Real-time high-resolution sparse voxelization with application to image-based modeling

    [TOG08] Articulated Mesh Animation from Multi-view Silhouettes   [data]

    [ECCV06] Carved visual hulls for image-based modeling 

    [3DPVT06] Visual shapes of silhouette sets 

    [CVIU04] Silhouette and stereo fusion for 3d object modeling 

    [CVPR03] Visual hull alignment and refinement across time: A 3d reconstruction algorithm combining shape-from-silhouette with stereo 

    [CGIT00] Image-based visual hulls 

    3. Multi-view Stereo

    [UIST16] Holoportation: Virtual 3d teleportation in real-time  

    [SIGGRAPH15] High-quality streamable free-viewpoint video  

    [TVCG10] A point-cloud-based multiview stereo algorithm for free-viewpoint video 

    [SIGGRAPH08] Markerless garment capture 

    [CGA07] Surface capture for performance-based animation 

    [TVC05] Scalable 3d video of dynamic scenes 

    4. Photometric Stereo

    [ICCV11] Shading-based dynamic shape refinement from multi-view video under general illumination

    [SIGGRAPH_Asia09] Dynamic shape capture using multi-view photometric stereo  

    5. Template based Approaches

    [TOG13] On-set performance capture of multiple actors with a stereo camera 

    [ECCV12] Full body performance capture under uncontrolled and varying illumination: A shading-based approach 

    [CVPR11] Markerless motion capture of interacting characters using multi-view image segmentation

    [SIGGRAPH_Asia10] Video-based reconstruction of animatable human characters 

    [SIGGRAPH08] Performance capture from sparse multi-view video 

    [CVPR09] Motion capture using joint skeleton tracking and surface estimation 

    [CVPR08] Robust fusion of dynamic shape and normal capture for high-quality reconstruction of time-varying geometry 

    6. Parametric Models

    [ICCV19] Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the Loop 

    [CVPR19] Expressive Body Capture: 3D Hands, Face, and Body from a Single Image  

    [CVPR18 Oral] Total Capture: A 3D Deformation Model for Tracking Faces, Hands, and Bodies  

    [CVPR18] End-to-end recovery of human shape and pose  [code and data] 

    [CVPR17] Unite the people: Closing the loop between 3d and 2d human representations [project page(code and data)] 

    [BMVC17] Indirect deep structured learning for 3D human body shape and pose prediction 

    [ECCV16] Keep it SMPL: Automatic estimation of 3d human pose and shape from a single image[code] 

    [SIGGRAPH_Asia15] SMPL: A skinned multi-person linear model 

    [TOG14] Mosh: Motion and shape capture from sparse markers  

    [CVPR10] Multilinear pose and body shape estimation of dressed subjects from image sets 

    [ICCV09] Estimating Human Shape and Pose from a Single Image 

    [CVPR07] Detailed human shape and pose from images 

    [SIGGRAPH05] SCAPE: shape completion and animation of people  

    [CVIU01] Tracking and modeling people in video sequences 

Single-view Reconstruction

    [ICCV19] Tex2Shape: Detailed Full Human Body Geometry From a Single Image 

    [ICCV19] PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitization 

    [ICCV19] Moulding Humans: Non-parametric 3D Human Shape Estimation from Single Images 

    [ICCV19] 3DPeople: Modeling the Geometry of Dressed Humans 

    [ICCV19] DeepHuman: 3D Human Reconstruction From a Single Image 

    [ICCV19] A Neural Network for Detailed Human Depth Estimation From a Single Image 

    [ICCV19] Delving Deep Into Hybrid Annotations for 3D Human Recovery in the Wild 

    [CVPR19 Oral] SiCloPe: Silhouette-Based Clothed People 

    [CVPR19 Oral] Convolutional Mesh Regression for Single-Image Human Shape Reconstruction 

    [CVPR19 Oral] Detailed Human Shape Estimation from a Single Image by Hierarchical mesh deformation 

    [CVPR19] Learning to Reconstruct People in Clothing from a Single RGB Camera  

    [CVPR19] SimulCap: Single-View Human Performance Capture with Cloth Simulation 

    [CVPR19] Photo Wake-Up: 3D Character Animation from a Single Photo  

    [TOG19] LiveCap: Real-time Human Performance Capture from Monocular Video  

    [CVPR18] End-to-end recovery of human shape and pose  [code and data] 

    [3DV18] Neural Body Fitting: Unifying Deep Learning and Model-Based Human Pose 

    [ICCV09] Estimating Human Shape and Pose from a Single Image 

4D-Scans

    [SIGGRAPH17] ClothCap: seamless 4D clothing capture and retargeting  

Datasets and Code

    MPI datasets and code

    BUFF

    Dynamic FAUST

    CVSSP3D

    Mixamo

    Human3.6M

    3DPeople

    THuman

    RenderPeople

    CLOTH3D(not released yet)

    Multi-Garment Net