Dataset Viewer
Auto-converted to Parquet Duplicate
The dataset viewer is not available for this split.
Parquet error: Scan size limit exceeded: attempted to read 1724037996 bytes, limit is 300000000 bytes Make sure that 1. the Parquet files contain a page index to enable random access without loading entire row groups2. otherwise use smaller row-group sizes when serializing the Parquet files
Error code:   TooBigContentError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

PiyoShogi Eval (paired, 4 devices)

ぴよ将棋の盤面認識モデルの評価用データセット。4機種の実機スクリーンショットを SFEN 単位で束ねた横持ち形式

対応機種

機種名 識別子 (devices 列) 画面解像度
iPhone 8 iPhone10,1 750 × 1334
iPhone XR iPhone11,8 828 × 1792
iPhone 15 iPhone15,4 1179 × 2556
iPad Air M3 iPad14,10 1640 × 2360

paired config

1 行 = 1 SFEN、4 機種分の画像を list で持つ。

Column Type 説明
sfen string SFEN形式の局面文字列
hash string SFENのSHA-256
type string 局面ソース種別(現状は全て existing = ぴよ将棋プリセット由来)
images Sequence(Image) 各機種のフルスクリーンショット (lossless WebP)
devices Sequence(string) images と同じ順序の機種識別子
bboxes Sequence(Sequence(int32)) 各機種の盤面領域 [x1, y1, x2, y2] (retina px)

行数: 1,000 SFEN × 4機種 = 4,000 画像。

from datasets import load_dataset
ds = load_dataset("ultemica/piyoshogi-eval", "paired", split="test")
row = ds[0]
for img, dev, bbox in zip(row["images"], row["devices"], row["bboxes"]):
    print(dev, img.size, bbox)

train/val との関係

このリポは train/val と leak なしの 1,000 SFEN を保持する。 学習データは ultemica/piyoshogiocr_paired / detector_paired config を参照。

キャプチャ方法

IPA-Patch/PiyoShot を用いて、指定SFENを ぴよ将棋アプリに描画し、画面全体を自動でスクリーンショット。

ライセンス

MIT

Downloads last month
132