3DGS-Mesh Envs: 3D Digital Twin Environments for Robotics Benchmarking

3DGS-Mesh digital twin environments are released for the HUGE-Bench robotics benchmark. The dataset contains 3D assets required for rendering and collision-aware evaluation, including point clouds and simplified meshes. It was authored by yu781986168 and last updated on 2026-06-24.

Use Cases

• Rendering synthetic training data for robotics agents based on the provided 3D assets.
• Conducting collision-aware evaluation of robot navigation policies within the digital twin environments.
• Benchmarking 3D scene understanding and reconstruction algorithms using the point cloud and mesh data.
• Training and evaluating visual SLAM (Simultaneous Localization and Mapping) systems in simulated environments.

Strengths

• Provides 3D assets specifically formatted for the HUGE-Bench robotics benchmark.
• Includes both point cloud (.ply) and simplified mesh (.obj) representations for each environment.
• Data is structured with separate train, test_seen, and test_unseen splits for benchmark tasks.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and dataset size are unknown, which may limit suitability assessment.

Provenance

Source

huggingface

Freshness

Last updated 2026-06-24 04:38:07; freshness should be verified.