FENiT Data: Chessboard Images for Digitization and FEN Generation

FENiT Data is a collection of chessboard images used for training and evaluating a YOLO model for automatic digitization and FEN generation. The dataset contains 3.1 GB of images captured under various lighting, angles, and piece positioning conditions. It was authored by Theo Deannata Harjanto and last updated on April 14, 2026.

Use Cases

• Train object detection models for chess pieces based on captured images.
• Evaluate model performance under varied lighting conditions mentioned in the description.
• Generate FEN notation from board states using computer vision.
• Benchmark digitization algorithms against a variety of angles and piece positioning.

Strengths

• Dataset size is 3.1 GB, indicating a substantial collection of images.
• Images are captured under varied lighting, angles, and piece positioning, likely enhancing model robustness.
• License is CC0-1.0, allowing for unrestricted use.

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

Theo Deannata Harjanto

Collection Method

Derived from a mobile application project for automatic chessboard digitization and FEN generation using computer vision.

Freshness

Last updated 2026-04-14 16:18:43; freshness should be verified.