3维人体建模历史最全算法、论文、数据等资源整理分享_3d人体模型 公开数据集

    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。

 本资源对3维人体建模相关的各种算法进行了分类梳理，对其中的算法、论文、数据资源等进行整理，分享给需要的朋友。

 

    资源整理自网络，资源获取见源地址：https://github.com/3DFaceBody/awesome-3dbody-papers

 

目录

论文资源列表

 

Body Model 

    SCAPE: Shape Completion and Animation of People. SIGGRAPH, 2005. 

 

    SMPL: A Skinned Multi-Person Linear Model. SIGGRAPH Asia, 2015.  

 

    Expressive Body Capture: 3D Hands, Face, and Body from a Single Image. CVPR, 2019.  

 

    SoftSMPL: Data-driven Modeling of Nonlinear Soft-tissue Dynamics for Parametric Humans. Eurographics, 2020. 

 

    STAR: Sparse Trained Articulated Human Body Regressor. ECCV, 2020.  

 

    BLSM: A Bone-Level Skinned Model of the Human Mesh. ECCV, 2020. 

 

    Joint Optimization for Multi-Person Shape Models from Markerless 3D-Scans. ECCV, 2020. 

 

    GHUM & GHUML: Generative 3D Human Shape and Articulated Pose Models. CVPR (Oral), 2020. 

 

Body Pose 

    MotioNet: 3D Human Motion Reconstruction from Monocular Video with Skeleton Consistency. ToG, 2020.  

 

    VNect: real-time 3D human pose estimation with a single RGB camera. SIGGRAPH Asia, 2017.  

 

    XNect: Real-time Multi-person 3D Human Pose Estimation with a Single RGB Camera. SIGGRAPH, 2020.  

 

    PhysCap: Physically Plausible Monocular 3D Motion Capture in Real Time. SIGGRAPH Asia, 2020. 

 

    Unsupervised 3D Human Pose Representation with Viewpoint and Pose Disentanglement. ECCV, 2020. 

 

    Cascaded Deep Monocular 3D Human Pose Estimation with Evolutionary Training Data. CVPR, 2020. 

 

    End-to-End Estimation of Multi-Person 3D Poses from Multiple Cameras. ECCV (Oral), 2020.

 

    Learnable Triangulation of Human Pose. ICCV (Oral), 2019. 

 

    Compressed Volumetric Heatmaps for Multi-Person 3D Pose Estimation. CVPR, 2020. 

 

    Multi-person 3D Pose Estimation in Crowded Scenes Based on Multi-View Geometry. arXiv, 2020. 

 

    PI-Net: Pose Interacting Network for Multi-Person Monocular 3D Pose Estimation. WACV, 2021.

 

    PandaNet: Anchor-Based Single-Shot Multi-Person 3D Pose Estimation. ArXiv, 2021.

 

    SMAP: Single-Shot Multi-Person Absolute 3D Pose Estimation. ECCV, 2020.  

 

    PI-Net: Pose Interacting Network for Multi-Person Monocular 3D Pose Estimation. WACV, 2021.

 

    Temporal Smoothing for 3D Human Pose Estimation and Localization for Occluded People. ArXiv, 2020. 

 

    MeTRAbs: Metric-Scale Truncation-Robust Heatmaps for Absolute 3D Human Pose Estimation. T-BIOM, 2020. 

 

    PCLs: Geometry-aware Neural Reconstruction of 3D Pose with Perspective Crop Layers. ArXiv, 2020.

 

    CanonPose: Self-Supervised Monocular 3D Human Pose Estimation in the Wild. ArXiv, 2020.

 

    DOPE: Distillation Of Part Experts for whole-body 3D pose estimation in the wild. ECCV, 2020. 

 

    MocapNET: Ensemble of SNN Encoders for 3D Human Pose Estimation in RGB Images. BMVC, 2019. 

 

    Invariant Teacher and Equivariant Student for Unsupervised 3D Human Pose Estimation. AIII, 2021. 

 

    Residual Pose: A Decoupled Approach for Depth-based 3D Human Pose Estimation. IROS, 2020. 

 

    PoP-Net: Pose over Parts Network for Multi-Person 3D Pose Estimation from a Depth Image. ArXiv, 2020. 

 

Naked Body Mesh 

    Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Image. ECCV, 2016.  

 

    Learning to Estimate 3D Human Pose and Shape from a Single Color Image. CVPR, 2018. 

 

    Neural Body Fitting: Unifying Deep Learning and Model Based Human Pose and Shape Estimation. 3DV (Oral), 2018. 

 

    Appearance Consensus Driven Self-Supervised Human Mesh Recovery. ECCV (Oral), 2020.  

 

    Delving Deep Into Hybrid Annotations for 3D Human Recovery in the Wild. ICCV, 2019.  

 

    Learning 3D Human Shape and Pose from Dense Body Parts. ArXiv, 2019.  

 

    Full-Body Awareness from Partial Observations. ECCV, 2020.  

 

    3D Multi-bodies: Fitting Sets of Plausible 3D Human Models to Ambiguous Image Data. NeurIPS, 2020.

 

    Parametric Shape Estimation of Human Body under Wide Clothing. ACM MM, 2020. 

 

    Human Body Model Fitting by Learned Gradient Descent. ECCV, 2020. 

 

    End-to-end Recovery of Human Shape and Pose. CVPR, 2018.  

 

    Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the Loop. ICCV, 2019.  

 

    3D Human Mesh Regression with Dense Correspondence. CVPR, 2020. 

 

    Hierarchical Kinematic Human Mesh Recovery. ECCV, 2020. 

 

    I2L-MeshNet: Image-to-Lixel Prediction Network for Accurate 3D Human Pose and Mesh Estimation from a Single RGB Image. ECCV, 2020. 

 

    PoseLifter: Absolute 3D Human Pose Lifting Network from a Single Noisy 2D Human Pose. ArXiv, 2020. 

 

    Pose2Mesh: Graph Convolutional Network for 3D Human Pose and Mesh Recovery from a 2D Human Pose. ECCV, 2020. 

 

    PoseNet3D: Learning Temporally Consistent 3D Human Pose via Knowledge Distillation. 3DV, 2020.

 

    Human Mesh Recovery from Monocular Images via a Skeleton-disentangled Representation. ICCV, 2019. 

 

    Learning 3D Human Shape and Pose from Dense Body Parts. TPAMI, 2020.  

 

    Exemplar Fine-Tuning for 3D Human Pose Fitting Towards In-the-Wild 3D Human Pose Estimation. ArXiv, 2020. 

 

    HybrIK: A Hybrid Analytical-Neural Inverse Kinematics Solution for 3D Human Pose and Shape Estimation. ArXiv, 2020. 

 

    Chasing the Tail in Monocular 3D Human Reconstruction with Prototype Memory. ArXiv, 2020.

 

    Beyond Weak Perspective for Monocular 3D Human Pose Estimation. ArXiv, 2020.

 

    CenterHMR: a Bottom-up Single-shot Method for Multi-person 3D Mesh Recovery from a Single Image. ArXiv, 2020. 

 

    Full-body motion capture for multiple closely interacting persons. CVM, 2020.

 

    Learning 3D Human Dynamics from Video. CVPR, 2019.  

 

    VIBE: Video Inference for Human Body Pose and Shape Estimation. CVPR, 2020. 

 

    3D Human Motion Estimation via Motion Compression and Refinement. ACCV (Oral), 2020.  

 

    Beyond Static Features for Temporally Consistent 3D Human Pose and Shape from a Video. ArXiv, 2020.

 

    End-to-End Human Pose and Mesh Reconstruction with Transformers. ArXiv, 2020.

 

    Human Mesh Recovery from Multiple Shots. ArXiv, 2020. 

 

    Total Capture: A 3D Deformation Model for Tracking Faces, Hands, and Bodies. CVPR (Oral), 2018. 

 

    Monocular Total Capture: Posing Face, Body and Hands in the Wild. CVPR (Oral), 2019.  

 

    Expressive Body Capture: 3D Hands, Face, and Body from a Single Image. CVPR, 2019.  

 

    FrankMocap: A Fast Monocular 3D Hand and Body Motion Capture by Regression and Integration. ArXiv, 2020.  

 

    Monocular Expressive Body Regression through Body-Driven Attention. ECCV, 2020.  

 

    NeuralAnnot: Neural Annotator for in-the-wild Expressive 3D Human Pose and Mesh Training Sets. ArXiv, 2020. 

 

    Pose2Pose: 3D Positional Pose-Guided 3D Rotational Pose Prediction for Expressive 3D Human Pose and Mesh Estimation. ArXiv, 2020. 

 

    Monocular Real-time Full Body Capture with Inter-part Correlations. ArXiv, 2020.

 

 Clothed Body Mesh

    LiveCap: Real-time Human Performance Capture from Monocular Video. SIGGRAPH, 2019. 

 

    DeepCap: Monocular Human Performance Capture Using Weak Supervision. CVPR (Oral), 2020. 

 

    MonoClothCap: Towards Temporally Coherent Clothing Capture from Monocular RGB Video. 3DV, 2020.

 

    MulayCap: Multi-layer Human Performance Capture Using A Monocular Video Camera. TVCG, 2020. 

 

    Deep Physics-aware Inference of Cloth Deformation for Monocular Human Performance Capture. ArXiv, 2020.

 

    DoubleFusion: Real-time Capture of Human Performance with Inner Body Shape from a Depth Sensor. CVPR (Oral), 2018.  

 

    SimulCap : Single-View Human Performance Capture with Cloth Simulation. CVPR, 2019. 

 

    Robust 3D Self-portraits in Seconds. CVPR (Oral), 2020. 

 

    RobustFusion: Human Volumetric Capture with Data-driven Visual Cues using a RGBD Camera. ECCV, 2020.

 

    NormalGAN: Learning Detailed 3D Human from a Single RGB-D Image. ECCV, 2020. 

 

    TexMesh: Reconstructing Detailed Human Texture and Geometry from RGB-D Video. ECCV, 2020. 

 

    Fast Generation of Realistic Virtual Humans. VRST, 2017. 

 

    Realistic Virtual Humans from Smartphone Videos. VRST, 2020. 

 

    Video Based Reconstruction of 3D People Models. CVPR, 2018. 

 

    Learning to Reconstruct People in Clothing from a Single RGB Camera. CVPR, 2019.  

 

    SiCloPe: Silhouette-Based Clothed People. CVPR, 2019.

 

    Tex2Shape: Detailed Full Human Body Geometry from a Single Image. ICCV, 2019.  

 

    Multi-Garment Net: Learning to Dress 3D People from Images. ICCV, 2019. 

 

    SIZER: A Dataset and Model for Parsing 3D Clothing and Learning Size Sensitive 3D Clothing. ECCV (Oral), 2020.  

 

    PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitization. ICCV, 2019.  

 

    PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization. CVPR (Oral), 2020.  

 

    Geo-PIFu: Geometry and Pixel Aligned Implicit Functions for Single-view Human Reconstruction. ArXiv, 2020.

 

    ARCH: Animatable Reconstruction of Clothed Humans. CVPR, 2020.

 

    S3: Neural Shape, Skeleton, and Skinning Fields for 3D Human Modeling. ArXiv, 2021.

 

    Detailed Human Avatars from Monocular Video. 3DV, 2018. 

 

    Monocular Real-Time Volumetric Performance Capture. ECCV, 2020.  

 

    Implicit Functions in Feature Space for 3D Shape Reconstruction and Completion. CVPR, 2020.  

 

    Combining Implicit Function Learning and Parametric Models for 3D Human Reconstruction. ECCV (Oral), 2020.  

 

    PaMIR: Parametric Model-Conditioned Implicit Representation for Image-based Human Reconstruction. TPAMI, 2020. 

 

    RIN: Textured Human Model Recovery and Imitation with a Single Image. ArXiv, 2020.

 

    3D Human Avatar Digitization from a Single Image. VRCAI, 2019.

 

    Reconstructing NBA Players. ECCV, 2020.  

 

Human Depth Estimation

    Learning the Depths of Moving People by Watching Frozen People. CVPR, 2019.  

 

    A Neural Network for Detailed Human Depth Estimation from a Single Image. ICCV, 2019. 

 

    Self-Supervised Human Depth Estimation from Monocular Videos. CVPR, 2020. 

 

    DressNet: High Fidelity Depth Estimation of Dressed Humans from a Single View Image. ArXiv, 2021.

 

Human Motion

    3D Semantic Trajectory Reconstruction from 3D Pixel Continuum. CVPR, 2018. 

 

    Convolutional Autoencoders for Human Motion Infilling. 3DV, 2020.

 

    Predicting 3D Human Dynamics from Video. ICCV, 2019.  

 

    Long-term Human Motion Prediction with Scene Context. ECCV (Oral), 2020.  

 

    Adversarial Refinement Network for Human Motion Prediction. ACCV, 2020.

 

    Synthesizing Long-Term 3D Human Motion and Interaction in 3D. ArXiv, 2020. 

 

    GlocalNet: Class-aware Long-term Human Motion Synthesis. MACV, 2021.

 

    Learn to Dance with AIST++: Music Conditioned 3D Dance Generation. ArXiv, 2021. 

 

    A Causal Convolutional Neural Network for Motion Modeling and Synthesis. ArXiv, 2021.

 

Human-Object Interaction

    Perceiving 3D Human-Object Spatial Arrangements from a Single Image in the Wild. ECCV, 2020.  

 

    Resolving 3D Human Pose Ambiguities with 3D Scene Constraints. ICCV, 2019.  

 

    GRAB: A Dataset of Whole-Body Human Grasping of Objects. ECCV, 2020.  

 

    Populating 3D Scenes by Learning Human-Scene Interaction. ArXiv, 2020. 

 

    Holistic 3D Human and Scene Mesh Estimation from Single View Images. ArXiv, 2020.

 

Animation

    Predicting Animation Skeletons for 3D Articulated Models via Volumetric Nets. 3DV (Oral), 2019.  

 

    RigNet: Neural Rigging for Articulated Characters. SIGGRAPH, 2020.  

 

    Skeleton-Aware Networks for Deep Motion Retargeting. SIGGRAPH, 2020.  

 

    Motion Retargetting based on Dilated Convolutions and Skeleton-specific Loss Functions. Eurographics, 2020.  

 

    DeePSD: Automatic Deep Skinning And Pose Space Deformation For 3D Garment Animation. ArXiv, 2020.

 

    UniCon: Universal Neural Controller For Physics-based Character Motion. ArXiv, 2020. 

 

Cloth/Try-On

    DeepWrinkles: Accurate and Realistic Clothing Modeling. ECCV (Oral), 2018.

 

    Deep Fashion3D: A Dataset and Benchmark for 3D Garment Reconstruction from Single-view Images. ECCV (Oral), 2020. 

 

    TailorNet: Predicting Clothing in 3D as a Function of Human Pose, Shape and Garment Style. CVPR (Oral), 2020.  

 

    Learning-Based Animation of Clothing for Virtual Try-On. Eurographics, 2019. 

 

    Physically Based Neural Simulator for Garment Animation. ArXiv, 2020.

 

    DeepCloth: Neural Garment Representation for Shape and Style Editing. ArXiv, 2020.

 

    Example-based Real-time Clothing Synthesis for Virtual Agents. ArXiv, 2021.

 

    BCNet: Learning Body and Cloth Shape from a Single Image. ECCV, 2020. 

 

    Fully Convolutional Graph Neural Networks for Parametric Virtual Try-On. SCA, 2020. 

 

    Neural 3D Clothes Retargeting from a Single Image. ArXiv, 2021.

 

Neural Rendering

    Neural3D: Light-weight Neural Portrait Scanning via Context-aware Correspondence Learning. ACM MM, 2020.

 

    Multi-view Neural Human Rendering. CVPR, 2020.  

 

    ANR: Articulated Neural Rendering for Virtual Avatars. ArXiv, 2020. 

 

    SMPLpix: Neural Avatars from 3D Human Models. WACV, 2020.  

 

    Vid2Actor: Free-viewpoint Animatable Person Synthesis from Video in the Wild. ArXiv, 2020. 

 

    Neural Body: Implicit Neural Representations with Structured Latent Codes for Novel View Synthesis of Dynamic Humans. ArXiv, 2020. 

 

Dataset

    3DPW: Recovering Accurate 3D Human Pose in The Wild Using IMUs and a Moving Camera. ECCV, 2018. 

 

    AMASS: Archive of Motion Capture as Surface Shapes. ICCV, 2019.  

 

    3DBodyTex: Textured 3D Body Dataset. 3DV, 2018. 

 

    Motion Capture from Internet Videos. ECCV, 2020.  

 

    Full-Body Awareness from Partial Observations. ECCV, 2020.  

 

    HUMBI: A Large Multiview Dataset of Human Body Expressions. ECVPR, 2020.  

 

    SMPLy Benchmarking 3D Human Pose Estimation in the Wild. 3DV (Oral), 2020.