CLARU
nepfaffapache-2.0

SceneSmith Preprocessed Data

Preprocessed 3D assets for SceneSmith, including articulated objects converted to Drake SDFormat and CLIP embeddings for PBR materials, enabling VLM-agent-based generation of physically realistic indoor scenes.

Downloads196

Why This Matters for Physical AI

This dataset enables physically realistic scene generation for robot simulation by providing articulated object models with accurate collision geometry and physical properties, supporting the development of embodied AI agents that interact with complex household environments.

Technical Profile

Modalities
3d_meshespoint_cloudclip_embeddings
Environment
simulationindoor
Task Types
scene_generationmanipulation
Data Format
Drake SDFormat (SDF), glTF, OBJ
Annotation Types
clip_embeddingsphysical_propertiesarticulated_joints
License
apache-2.0
Part of the SceneSmith family

Community Signals

Top 50% by downloads

Access

View on HuggingFace

Need custom 3d_meshes data?

Claru builds purpose-built datasets for simulation applications with dense human annotations and quality assurance.

Request a Sample Pack

Related Datasets

Xperience-10M

A large-scale egocentric multimodal dataset of human experience containing 10 million interactions and 10,000 hours of synchronized first-person recordings with six video streams, audio, stereo depth, camera pose, hand mocap, full-body mocap, IMU, and hierarchical language annotations for embodied AI, robotics, and world modeling research.

rgbaudiodepthproprioception+3
2.3M downloadsother

TartanGround

A large-scale, multi-modal dataset designed to advance perception and autonomy of ground robots operating in diverse environments, collected across 63 photorealistic simulation environments with over 1.44 million samples from 878 trajectories.

rgbdepthsemantic_segmentationoptical_flow+5
42K downloads2025cc-by-4.0

ManiTwin-100K: Manipulation-Ready Digital Object Twins

A large-scale dataset of 100K manipulation-ready digital object twins with simulation-ready 3D meshes, physical properties, functional point annotations, grasp configurations, and language descriptions validated through physics-based simulation.

point_cloudlanguage
13K downloads2026apache-2.0

SandGO Dataset (v1.0)

A long-horizon multimodal dataset designed for embodied AI and robot learning, containing state-action sequences, multi-view RGB images, LiDAR point clouds, depth point clouds, and natural language instructions for training Vision-Language-Action models.

rgbdepthpoint_cloudproprioception+1
9K downloadsapache-2.0

FastUMI Pro – Multimodal Sample Dataset

Small-scale demonstration data from the FastUMI Pro multimodal sensing system, containing dozens of sample trajectories with synchronized RGB images, point clouds, pose tracking, and haptic feedback. Full-scale datasets are available upon request.

rgbpoint_cloudproprioceptiontactile
8K downloadsother

Physical AI NuRec Dataset

A collection of 3D reconstructed indoor environments in USD/USDZ format designed for Isaac Sim simulation, including meshes, occupancy maps, and neural rendering components for robotic applications.

rgbdepthpoint_cloud
660 downloads2025cc-by-4.0