fullseye

fly_emd_response — FLYVISION motion op

• データ種: signal × signal → signal
• 呼び出し: import fullseye as fs; fs.ledger.fly_emd_response(signal_a, signal_b, tau_s=0.05, dt_s=0.001) (実装を直接呼ぶなら import flyvision; flyvision.fly_emd_response(signal_a, signal_b, tau_s=0.05, dt_s=0.001)、台帳から引くなら opsflyvision.get("fly_emd_response"))

使い方

Hassenstein-Reichardt correlator between two adjacent ommatidial signals.

The opponent elementary motion detector::

R(t) = LP(a)(t) * b(t) - a(t) * LP(b)(t)

where LP is a first-order low-pass of time constant tau_s. Delaying one channel and multiplying it against the other, then subtracting the mirror pair, gives a signal whose sign is the direction of motion and whose steady-state mean encodes the temporal frequency.

signal_a / signal_b: the two 1-D input time series (same length). tau_s: the low-pass time constant, seconds. dt_s: the sample interval, seconds.

Returns a 1-D float64 array R(t) of the same length.

Ground truth: for two sinusoids of temporal frequency f (omega = 2 pi f) with a spatial phase psi = 2 pi dphi/lambda between them, the steady-state mean is R_bar = dI^2 * sin(psi) * omega tau/(1 + (omega tau)^2) (to 5%), which is zero at lambda = 2 dphi (psi = pi) and maximal at f = 1/(2 pi tau) independent of lambda (pinned in the tests).

Raises ValueError: a non-1-D / empty / too-short (< 2) / non-finite signal_a or signal_b, mismatched lengths, a signal over :data:MAX_SIGNAL_POINTS, and a non-positive tau_s / dt_s.

参考(サンプルデータ・文献)

• サンプルデータ カタログ(DL URL / ライセンス) — 2-D は skimage.data(BSD/public)+ 合成、3-D は実データ源(Stanford/PDS 等)の DL URL。
• 演算子の来歴・参考文献 — この op 族の元になった研究/手法の出典。
• アルゴリズムの正典(著者・年)と用途は上記ファミリ使い方ガイドに記載。

実行できる例(この op を実際に呼ぶ検証済みサンプル)

• poc_fly_vision — py -3.11 examples/poc_fly_vision.py

型が繋がる次の op(signal を入力に取れる)

fly_lgmd_eta · fly_tau_from_expansion · fly_dsi

同カテゴリ(motion)

—

---

Provenance: flyvision.py — FLYVISION operator registry. この per-op ノートは tools/opdocs.py md が自動生成(手編集しない)。

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

This site is open source. Improve this page.