S2O: Static to Openable Enhancement for Articulated 3D Objects

3dlg-hcvc provides mesh, point cloud, and metadata for two datasets used in the S2O research project. The PM-Openable subset contains 648 openable objects from PartNet-Mobility, with a train/val/test split of 460/95/93 objects. The Articulated Container Dataset (ACD) contains openable container objects sourced from HSSD.

Use Cases

• Train models for static-to-openable 3D object enhancement based on the provided mesh and point cloud data.
• Benchmark articulated object generation and reconstruction algorithms using the PM-Openable subset.
• Develop models for container object manipulation using the Articulated Container Dataset (ACD).

Strengths

• Provides data for 648 openable 3D objects in the PM-Openable subset.
• Includes a defined train/val/test split of 460/95/93 objects for structured evaluation.
• Offers multiple data representations including mesh, point cloud, and metadata.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the Articulated Container Dataset (ACD) is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

3dlg-hcvc via Hugging Face.

Collection Method

Subsets derived from PartNet-Mobility and HSSD datasets.

Freshness

Last updated 2025-05-15 20:48:08; freshness should be verified.