SynCom (Synthetic Composition) is a synthetic dataset for realistic 3D object-scene composition. The dataset, authored by yGaoJiany, provides controlled object placement, known camera parameters, and rendered ground truth for composed scenes. It was last updated on April 29, 2026.
Use Cases
- Evaluate object-scene appearance matching based on rendered ground truth.
- Train models for physically plausible shadow generation based on controlled object placement.
- Benchmark 3D reconstruction algorithms based on known camera parameters.
- Develop models for multi-view image set composition tasks.
- Research synthetic data generation for 3D computer vision tasks.
Strengths
- Provides controlled object placement, a feature mentioned in the description.
- Includes known camera parameters, a feature mentioned in the description.
- Contains rendered ground truth for composed scenes, a feature mentioned in the description.
- Last updated on April 29, 2026.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
- Description metadata is limited; actual data quality requires manual inspection after download.
Provenance
- Source
- yGaoJiany
- Collection Method
- Synthetic generation for 3D object-scene composition.
- Freshness
- Last updated 2026-04-29 03:34:09.