Semi-Structured: Screw Gear 02
SUMMARY
Photorealistic RGB renders of industrial parts in semi-structured scenes, generated with Extreme Domain Randomization (EDR) for sim-to-real transfer, with pixel-accurate instance masks. Access this dataset on Kaggle: Semi-Structured Screw Gear 02
Samples
About the Dataset
A synthetic dataset for instance segmentation of industrial metallic parts in semi-structured scenes, targeting applications in robotic manipulation, sim-to-real transfer, and object detection.
Each scene features industrial components placed in semi-structured arrangements and rendered under Extreme Domain Randomization (EDR) — aggressive variation of lighting, materials, and backgrounds — yielding photorealistic RGB images with per-instance segmentation masks that transfer well from simulation to real-world inference.
Specifications
| Kaggle link | Semi-Structured Screw Gear 02 |
| Total images | 5,000 |
| Train-Test-Val Split (percentage) | 70% – 10% – 20% |
| Image resolution | 848 × 480 |
| Image format | PNG (8-bit sRGB) |
| Annotation format | COCO JSON, instance masks as encoded RLE |
| Object classes | 1 (gear, category_id=6, supercategory=object) |
| License | CC BY-NC-SA 4.0 |