Synlayers Data: Layered Image Samples for Training

SynLayers training data contains layered image samples for computer vision tasks. Each row includes a base image, a composited whole image, separate layer images with captions and bounding boxes, and sanitized metadata. The dataset was created by SynLayers and was last updated on 2026-05-18.

Use Cases

• Train image generation models based on layered image composition data.
• Develop object detection models based on bounding-box caption annotations.
• Benchmark scene understanding algorithms based on the separation of image layers.
• Create synthetic training datasets for computer vision tasks based on the provided layer structure.

Strengths

• Includes bounding-box caption training annotations in a dedicated JSON file.
• Each sample contains multiple aligned data types: base PNG, composited PNG, layer PNGs, captions, and bounding boxes.
• Original metadata is included in a sanitized payload.

Limitations

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

Provenance

Source

SynLayers

Freshness

Last updated 2026-05-18 14:02:48; freshness should be verified.