Human-Authored 3D Interior Scenes with 211 Synthetic Environments

The Habitat Synthetic Scenes Dataset (HSSD) contains 211 human-authored 3D scenes and over 18,000 models of real-world objects. It is designed to mirror real interiors more closely than prior synthetic datasets for consumption in the Habitat simulation platform.

Use Cases

• Train embodied AI agents for navigation and interaction tasks within the 211 synthetic 3D interior scenes.
• Benchmark computer vision models for 3D scene understanding using the dataset's diverse object models.
• Develop robotics simulation environments by leveraging the Habitat-ready 3D scene assets.

Strengths

• Contains 211 distinct 3D interior scenes authored by humans.
• Includes over 18,000 models of real-world objects for scene composition.
• Designed to more closely mirror real-world scenes compared to prior synthetic datasets.

Limitations

• Dataset size, file formats, and specific column structure are not detailed in the provided input.
• The dataset is synthetic, which may introduce a domain gap compared to real-world sensor data.
• The temporal and geographic coverage of the represented scenes is unspecified.

Provenance

Source

hssd on Hugging Face.

Collection Method

Human-authored synthetic 3D scenes.

Freshness

Last updated on 2025-02-14.