video_summary_keyframes — VIDEOCUBE summary op

• Data kinds: video → indices

• Call: import fullseye as fs; fs.ledger.video_summary_keyframes(video, k: 'int' = 5, min_gap: 'int | None' = None, bins: 'int' = 32) -> 'np.ndarray' (to call the implementation directly, import videocube; videocube.video_summary_keyframes(video, k: 'int' = 5, min_gap: 'int | None' = None, bins: 'int' = 32) -> 'np.ndarray'; from the registry, opsvideocube.get("video_summary_keyframes"))

Usage

> This operator's description has not been translated yet. The original text follows as it is.

代表フレームの添字 `(k,) int64、昇順(indices`)。

フレームごとの得点 = 動きの量(|I_t − I_{t−1}| の平均、[0, 1] に正規化)+ 場面の変化(ヒストグラムの

χ² 距離、正規化)。得点の高い順に、既に選んだフレームから `min_gap`(既定 T // (2k))以上離れたものを

貪欲に `k` 枚選ぶ。T < k なら全フレーム。学習なしの「何かが起きたフレーム」であって、意味の要約ではない。

>>> idx = video_summary_keyframes(clip, k=4)

>>> thumbnails = clip[idx]

Detailed usage guide

• videocube family guide

References (sample data, literature)

• Sample-data catalog (download URLs / licences) — 2-D uses skimage.data (BSD/public domain) plus synthetic images; 3-D lists download URLs for real data sources (Stanford, PDS, …).

• Operator provenance and references — the sources of the research/methods this op family came from.

• The canonical algorithm (author, year) and its uses are named in the family usage guide above.

Runnable examples (verified samples that actually call this op)

• poc_video_cube — py -3.11 examples/poc_video_cube.py

Ops the type connects to (they accept indices as input)

—

Same category (summary)

—

*Provenance: videocube.py — VIDEOCUBE operator registry. This per-op note is generated by tools/opdocs.py md (do not hand-edit).*

© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.