Dataset Details
Explore our diverse range of datasets and products below.
Our most popular & used dataset that started it all
Posed CLASSIC Dataset
Dataset Details
| No. of refined 3D models | 3350 single models 170 group models |
| No. of unique individuals | 203 |
| Description | static scans full-body dressed everyday poses everyday clothing including items / objects |
| Available Extras | Basic Metadata Multi-View Images (partially) Learn more about our extras. |
Technical Specifications
| Geometry | 100k & 30k polygon version triangulated Renderpeople 1.0 scanner |
| Texture Resolution | 8k textures from the Renderpeople 1.0 scanner |
| Textures | Diffuse Map Normal Map Basic masks (no PBR) |
| Available File Formats | 3ds Max Maya Blender Cinema 4D SketchUp Rhino OBJ FBX |
Posed Classic Dataset References
Selected research projects utilizing the Posed Classic dataset:
2024-07-17 | Generalizable Human Gaussians for Sparse View Synthesis
2024-06-12 | Human 3Diffusion: Realistic Avatar Creation via Explicit 3D Consistent Diffusion Models
2024-06-11 | Instant 3D Human Avatar Generation using Image Diffusion Models
2024-03-15 | ANIM: Accurate Neural Implicit Model for Human Reconstruction from a single RGB-D image
2023-12-30 | SHARE: Single-view Human Adversarial REconstruction
2023-12-18 | Paint-it: Text-to-Texture Synthesis via Deep Convolutional Texture Map Optimization and Physically-Based Rendering
2023-10-04 | GETAvatar: Generative Textured Meshes for Animatable Human Avatars
2023-05-23 | NCHO: Unsupervised Learning for Neural 3D Composition of Humans and Objects
2022-12-14 | ECON: Explicit Clothed humans Optimized via Normal integration
2022-12-13 | Structured 3D Features for Reconstructing Controllable Avatars
2022-08-12 | PRIF: Primary Ray-based Implicit Function
2022-05-13 | Monocular Human Digitization via Implicit Re-projection Networks
2022-04-25 | Generalizable Neural Performer: Learning Robust Radiance Fields for Human Novel View Synthesis
2022-04-19 | Photorealistic Monocular 3D Reconstruction of Humans Wearing Clothing
2022-01-11 | gDNA: Towards Generative Detailed Neural Avatars
2021-04-29 | AGORA: Avatars in Geography Optimized for Regression Analysis
2020-12-21 | POSA: Populating 3D Scenes by Learning Human-Scene Interaction
2020-04-01 | PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization
2019-05-13 | PIFu: Pixel-Aligned Implicit Function for High-Resolution Clothed Human Digitization
View even more research publications featuring our products here.
Our newest, most diverse and fastest growing dataset
Posed NewGen Dataset
Dataset Details
| No. of refined 3D models | 2500 (growing weekly) |
| No. of unique individuals | 500 (view diversity stats here) |
| No. of total available scans | 14500 (e.g. available as Multi-View Images) |
| Description | static scans full-body dressed everyday poses everyday clothing including items / objects |
| Available Extras | Detailed Metadata (learn more) Multi-View Images Point Cloud Data UHQ RAW Meshes Body Measurements 3D Body Labels Additional Poses Isolated Object Scans Learn more about our extras. |
Technical Specifications
| Geometry | 50k polygons triangulated Renderpeople 2.0 scanner |
| Texture Resolution | crisp 4k textures based on 16k RAW downsampling Renderpeople 2.0 scanner |
| Included Textures | Basecolor Normals Roughness Specular (PBR Ready) |
| Available File Formats | 3ds Max Maya Blender Unreal Engine 5 Unity Rhino SketchUp Cinema 4D OBJ FBX GLB |
Our highest quality premium 3D models
Posed PLUS Dataset
Dataset Details
| No. of refined 3D models | 500 |
| No. of unique individuals | 100 |
| Description | static scans full-body dressed everyday poses everyday clothing no items / objects 4x color variations per model |
| Available Extras | Detailed Metadata (learn more) Multi-View Images Point Cloud Data UHQ RAW Meshes Body Measurements 3D Body Labels Additional Poses Isolated Object Scans Learn more about our extras. |
Technical Specifications
| Geometry | 300k (LOD0) & 100k (LOD1) polygon version triangulated Renderpeople 2.0 scanner |
| Texture Resolution | 8k textures based on 16k RAW downsampling from the Renderpeople 2.0 scanner |
| Textures | Basecolor (4x color variation) Normals Roughness Specular Metalness Separation Masks (PBR ready) |
| Available File Formats | 3ds Max Maya Blender Unreal Engine 5 Unity Cinema 4D OBJ FBX GLB |
Our most versatile and flexible dataset
Rigged Dataset
Dataset Details
| No. of refined 3D models | 793 single models |
| No. of unique individuals | 129 |
| Description | rigged scans available as A-Pose and T-Pose full-body dressed everyday clothing no items / objects |
| Available Extras | Basic Metadata Multi-View Images (partially) Learn more about our extras. |
Technical Specifications
| Geometry | 10-15k polygons retopologized quads Renderpeople 1.0 scanner |
| Texture Resolution | 8k resolution Renderpeople 1.0 scanner |
| Textures | Diffuse map Normal map Gloss map 2x masks for separation of major clothing items |
| Available File Formats | 3ds Max Maya Cinema 4D Unreal Engine Unity FBX |
Rigged Dataset References
Selected research projects utilizing the Rigged dataset:
2024-07-15 | ConTEXTure: Consistent Multiview Images to Texture
2024-03-15 | NECA: Neural Customizable Human Avatar
2024-01-22 | CloSe: A 3D Clothing Segmentation Dataset and Model
2023-12-30 | 3D Human Pose Perception from Egocentric Stereo Videos
2023-12-07 | PrimDiffusion: Volumetric Primitives Diffusion for 3D Human Generation
2023-10-20 | PACE: Human and Camera Motion Estimation from in-the-wild Videos
2023-08-10 | Flexible Isosurface Extraction for Gradient-Based Mesh Optimization
2023-03-22 | SHERF: Generalizable Human NeRF from a Single Image
2022-12-13 | Structured 3D Features for Reconstructing Controllable Avatars
2022-09-22 | GET3D: A Generative Model of High Quality 3D Textured Shapes Learned from Images
2022-08-18 | MvDeCor: Multi-view Dense Correspondence Learning for Fine-grained 3D Segmentation
2022-08-02 | UnrealEgo: A New Dataset for Robust Egocentric 3D Human Motion Capture
2022-04-06 | AUV-Net: Learning Aligned UV Maps for Texture Transfer and Synthesis
2021-12-23 | HSPACE: Synthetic Parametric Humans Animated in Complex Environments
2021-12-17 | PeopleSansPeople: A Synthetic Data Generator for Human-Centric CV
2021-03-29 | HumanGPS: Geodesic PreServing Feature for Dense Human Correspondences
2020-07-30 | SimPose: Effectively Learning DensePose and Surface Normals of People from Simulated Data
2020-07-28 | Monocular Real-Time Volumetric Performance Capture
View even more research publications featuring our products here.
Flexible animations based on motion capture technology
Motion Animated Dataset
Dataset Details
| No. of refined 3D models | 204 single 3D models |
| No. of unique individuals | 90 |
| No. of unique mocap animations | 51 (each animation is used on multiple models) |
| Description | animated 3D models motion capture data applied to skeletal rig all mocap recorded with XSens MVN full-body dressed everyday movements everyday clothing no items / objects |
| Available Extras | Basic Metadata Multi-View Images (partially) Learn more about our extras. |
| Complementary Products | - The corresponding rigged people on which the mocap animations have been applied - Posed Classic models of the same individuals as in this Animated dataset |
Technical Specifications
| Geometry | 10-15k polygons retopologized quads Renderpeople 1.0 scanner |
| Texture Resolution | 8k resolution Renderpeople 1.0 scanner |
| Textures | Diffuse map Normal map Gloss map 2x masks for separation of major clothing items |
| Available File Formats | 3ds Max Maya Cinema 4D Unreal Engine Unity FBX |
Hyper-realistic motion and expression
Motion 4D Dataset
Dataset Details
| No. of refined 3D models | 123 single models |
| No. of unique individuals | 34 |
| Description | volumetric video capture (by Volucap) geometry & texture sequences full-body dressed everyday movements everyday clothing including items / objects |
| Available Extras | Basic Metadata Learn more about our extras. |
| More information | https://4dpeople.com/ |
Technical Specifications
| Geometry | optimized 30k polygon geometry sequence |
| Framerate | 30 fps |
| Texture Resolution | 4k resolution |
| Included Textures | Diffuse map sequence |
| Available File Formats | 3ds Max Maya Cinema 4D Blender Alembic |
a great addition to our other datasets
Bodyshape Dataset
Dataset Details
| No. of unique individuals | 500 (view diversity stats here) - this is the same pool of individuals as in the "Posed NewGen" and "Posed Plus" datasets |
| No. of total available scans | 1000 (each individual with and without hairnet) |
| Description | static scans full-body A-Pose body shape accentuating clothes no items / objects |
| Available Extras | Detailed Metadata (learn more) Multi-View Images Point Cloud Data UHQ RAW Meshes Body Measurements 3D Body Labels Additional Poses (in everyday clothing) Learn more about our extras. |
Technical Specifications
| Geometry | 50k polygons triangulated Renderpeople 2.0 scanner |
| Texture Resolution | crisp 4k textures based on 16k RAW downsampling Renderpeople 2.0 scanner |
| Included Textures | Basecolor Normals Roughness Specular (PBR Ready) |
| Available File Formats | 3ds Max Maya Blender Unreal Engine 5 Unity Rhino SketchUp Cinema 4D OBJ FBX GLB |