mirror of
https://github.com/ruvnet/RuView
synced 2026-07-24 17:43:20 +00:00
research(R6.2.3): chest-centric placement — +26.9 pp coverage gain for vital-signs cogs (#726)
Direct follow-up from R6.1 (chest contributes 27.6% of CSI energy, 5x per-limb value, limbs are confound not signal). R6.2.3 re-runs R6.2's placement search with chest-only target zones (40x40 cm patches at expected chest positions) vs body-footprint zones (R6.2's default full-area definition). Headline result: | Configuration | Coverage | Placement | |----------------------------|---------:|----------------------------| | Body-centric (R6.2 default)| 49.3% | (4.25,0)-(0,3.25), 5.35 m | | CHEST-CENTRIC (R6.2.3 new) | 82.4% | (2.0,0)-(4.5,5), 5.59 m | Cross-eval: - Body-optimal on chest zones: 55.5% - Chest-targeting GAIN on chest: +26.9 pp - Chest-optimal on body zones: 40.3% (-9.0 pp loss) The two strategies are genuinely different. Same engine, different zones. Per-cog deployment recommendation surfaced: - --target-mode=body (default): cog-person-count, cog-pose, cog-presence - --target-mode=chest (new): cog-vital-signs, cog-breathing, cog-HR - --target-mode=extremity (future): gesture detection ~20 LOC change to R6.2 CLI. R14 vertical-specific: - V1 stress-responsive lighting: chest mode - V2 adaptive HVAC (presence+breathing): mixed - V3 attention-respecting conversation: chest mode R6.2.3 surfaces a per-cog config that empathic-appliance products need at install time. Why placements differ: when target ~ envelope width, envelope can cover it entirely; when target >> envelope, placement must compromise. 40 cm Fresnel envelope @ 5 m link comfortably covers 40 cm chest patches but must spread to cover 3 m^2 bed. Composes: - R6.1 motivated this tick - R6.2 / R6.2.1 / R6.2.2 -- orthogonal extensions - R14 V1/V3 should use chest mode - R12 PABS improves body-position-detection scenarios Honest scope: - Chest positions approximated - 2D still (3D chest-centric = R6.2.3.1 follow-up) - Single subject (multi-subject = union of chest envelopes) - Per-cog zone schema is deployment-time Coordination: ticks/tick-23.md, no PROGRESS.md edit.
This commit is contained in:
rUv
@@ -0,0 +1,195 @@
|
||||
#!/usr/bin/env python3
|
||||
"""R6.2.3 — Chest-centric target zones for placement search.
|
||||
|
||||
See docs/research/sota-2026-05-22/R6_2_3-chest-centric-placement.md.
|
||||
|
||||
R6.1 quantified that the chest contributes 27.6% of the total CSI
|
||||
energy from a standing human -- 5x any single limb. R15's gait /
|
||||
breathing / RCS primitives are all dominated by chest dynamics.
|
||||
|
||||
This tick re-runs R6.2's placement search with chest-only target zones
|
||||
instead of full-body zones, and asks:
|
||||
|
||||
Does the optimal placement change when we target chest specifically?
|
||||
How much coverage is gained by aiming at the chest envelope alone?
|
||||
|
||||
If the answer is "no change", placement-time chest centring is
|
||||
unnecessary. If the answer is "significant change", R6.2's CLI tool
|
||||
should learn pose-aware zone definitions.
|
||||
|
||||
Pure NumPy.
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import argparse
|
||||
import json
|
||||
from pathlib import Path
|
||||
import numpy as np
|
||||
|
||||
C = 2.998e8
|
||||
|
||||
|
||||
def wavelength_m(freq_ghz: float) -> float:
|
||||
return C / (freq_ghz * 1e9)
|
||||
|
||||
|
||||
def in_first_fresnel(x, y, tx, rx, wavelength):
|
||||
r1 = np.sqrt((x - tx[0])**2 + (y - tx[1])**2)
|
||||
r2 = np.sqrt((x - rx[0])**2 + (y - rx[1])**2)
|
||||
direct = np.linalg.norm(tx - rx)
|
||||
return (r1 + r2) <= (direct + wavelength / 2)
|
||||
|
||||
|
||||
def coverage(tx, rx, target_zones, wavelength, resolution=0.05):
|
||||
per_zone = {}
|
||||
total_pts, total_covered = 0, 0
|
||||
for name, x0, y0, w, h in target_zones:
|
||||
xs = np.arange(x0, x0 + w, resolution)
|
||||
ys = np.arange(y0, y0 + h, resolution)
|
||||
gx, gy = np.meshgrid(xs, ys)
|
||||
mask = in_first_fresnel(gx.ravel(), gy.ravel(), tx, rx, wavelength)
|
||||
n_pts = len(gx.ravel())
|
||||
per_zone[name] = {
|
||||
"area_m2": float(n_pts * resolution ** 2),
|
||||
"covered_m2": float(mask.sum() * resolution ** 2),
|
||||
"coverage_fraction": float(mask.mean()),
|
||||
}
|
||||
total_pts += n_pts
|
||||
total_covered += mask.sum()
|
||||
return {
|
||||
"total_coverage_fraction": float(total_covered / total_pts) if total_pts > 0 else 0,
|
||||
"per_zone": per_zone,
|
||||
}
|
||||
|
||||
|
||||
def candidate_positions(room_w, room_h, step):
|
||||
cands = []
|
||||
for x in np.arange(0, room_w + 0.001, step):
|
||||
cands.append(np.array([x, 0.0]))
|
||||
cands.append(np.array([x, room_h]))
|
||||
for y in np.arange(step, room_h, step):
|
||||
cands.append(np.array([0.0, y]))
|
||||
cands.append(np.array([room_w, y]))
|
||||
return cands
|
||||
|
||||
|
||||
def search(target_zones, room_w, room_h, freq_ghz, step):
|
||||
lam = wavelength_m(freq_ghz)
|
||||
cands = candidate_positions(room_w, room_h, step)
|
||||
best = {"score": -1, "tx": None, "rx": None, "per_zone": None}
|
||||
for i, tx in enumerate(cands):
|
||||
for j, rx in enumerate(cands):
|
||||
if j <= i: continue
|
||||
if np.linalg.norm(tx - rx) < 1.0: continue
|
||||
cov = coverage(tx, rx, target_zones, lam)
|
||||
if cov["total_coverage_fraction"] > best["score"]:
|
||||
best = {
|
||||
"score": cov["total_coverage_fraction"],
|
||||
"tx": tx.tolist(), "rx": rx.tolist(),
|
||||
"link_m": float(np.linalg.norm(tx - rx)),
|
||||
"per_zone": cov["per_zone"],
|
||||
}
|
||||
return best
|
||||
|
||||
|
||||
def main():
|
||||
parser = argparse.ArgumentParser()
|
||||
parser.add_argument("--out", default="examples/research-sota/r6_2_3_chest_centric_results.json")
|
||||
args = parser.parse_args()
|
||||
|
||||
room_w, room_h = 5.0, 5.0
|
||||
freq = 2.4
|
||||
step = 0.25
|
||||
|
||||
# === BODY-CENTRIC zones (R6.2 default) ===
|
||||
# Bed (full lying area), chair (full sitting area), desk (full sitting area)
|
||||
body_zones = [
|
||||
("bed", 1.5, 0.5, 2.0, 1.5),
|
||||
("chair", 3.5, 3.5, 0.8, 0.8),
|
||||
("desk", 0.2, 2.5, 1.0, 0.6),
|
||||
]
|
||||
|
||||
# === CHEST-CENTRIC zones (R6.2.3 new) ===
|
||||
# The chest is approximately the upper-torso 40x30 cm region of the body.
|
||||
# Bed lying: chest at (2.5, 1.0) ± 30 cm
|
||||
# Chair sitting: chest at (3.9, 3.9) ± 20 cm
|
||||
# Desk: chest at (0.7, 2.8) ± 20 cm
|
||||
chest_zones = [
|
||||
("bed_chest", 2.2, 0.8, 0.6, 0.4), # 60x40 cm chest patch
|
||||
("chair_chest", 3.7, 3.7, 0.4, 0.4), # 40x40 cm
|
||||
("desk_chest", 0.5, 2.7, 0.4, 0.2), # 40x20 cm
|
||||
]
|
||||
|
||||
print(f"Room: {room_w}x{room_h} m, freq {freq} GHz")
|
||||
print()
|
||||
|
||||
print("=== Body-centric placement search ===")
|
||||
best_body = search(body_zones, room_w, room_h, freq, step)
|
||||
print(f" Best Tx: {best_body['tx']}, Rx: {best_body['rx']}")
|
||||
print(f" Link length: {best_body['link_m']:.2f} m")
|
||||
print(f" Total body-area coverage: {best_body['score']*100:.1f}%")
|
||||
print()
|
||||
|
||||
print("=== Chest-centric placement search ===")
|
||||
best_chest = search(chest_zones, room_w, room_h, freq, step)
|
||||
print(f" Best Tx: {best_chest['tx']}, Rx: {best_chest['rx']}")
|
||||
print(f" Link length: {best_chest['link_m']:.2f} m")
|
||||
print(f" Total chest-area coverage: {best_chest['score']*100:.1f}%")
|
||||
print()
|
||||
|
||||
# Cross-eval: how does the body-optimal placement perform on chest zones?
|
||||
lam = wavelength_m(freq)
|
||||
body_pl_on_chest = coverage(
|
||||
np.array(best_body["tx"]), np.array(best_body["rx"]), chest_zones, lam
|
||||
)
|
||||
chest_pl_on_body = coverage(
|
||||
np.array(best_chest["tx"]), np.array(best_chest["rx"]), body_zones, lam
|
||||
)
|
||||
|
||||
print("=== Cross-evaluation ===")
|
||||
print(f" Body-optimal placement on CHEST zones: {body_pl_on_chest['total_coverage_fraction']*100:.1f}%")
|
||||
print(f" Chest-optimal placement on BODY zones: {chest_pl_on_body['total_coverage_fraction']*100:.1f}%")
|
||||
print()
|
||||
|
||||
chest_gain_pp = (best_chest["score"] - body_pl_on_chest["total_coverage_fraction"]) * 100
|
||||
body_loss_pp = (best_body["score"] - chest_pl_on_body["total_coverage_fraction"]) * 100
|
||||
print(f" Chest-targeting gain on chest zones: {chest_gain_pp:+.1f} pp")
|
||||
print(f" Body-loss when using chest-optimal: {body_loss_pp:+.1f} pp")
|
||||
print()
|
||||
|
||||
# Verdict
|
||||
if abs(np.array(best_chest["tx"]) - np.array(best_body["tx"])).sum() < 0.6 and \
|
||||
abs(np.array(best_chest["rx"]) - np.array(best_body["rx"])).sum() < 0.6:
|
||||
verdict = "PLACEMENT STABLE: chest-centric search produces nearly the same optimal placement as body-centric. R6.2.3 is unnecessary at the placement-time level; chest-centric matters in the DSP pipeline (vital_signs.rs limb-mask), not the geometry."
|
||||
elif chest_gain_pp > 10:
|
||||
verdict = "CHEST-CENTRIC WINS: significant placement-strategy change. R6.2.3 should be a CLI option."
|
||||
else:
|
||||
verdict = "MIXED: chest and body placements differ but coverage gain is moderate. Documentation says use chest-centric for vital-signs cogs, body-centric for pose / count cogs."
|
||||
print(f"VERDICT: {verdict}")
|
||||
print()
|
||||
|
||||
out = {
|
||||
"room": {"width_m": room_w, "height_m": room_h},
|
||||
"freq_ghz": freq,
|
||||
"body_zones": [{"name": n, "x": x0, "y": y0, "w": w, "h": h}
|
||||
for n, x0, y0, w, h in body_zones],
|
||||
"chest_zones": [{"name": n, "x": x0, "y": y0, "w": w, "h": h}
|
||||
for n, x0, y0, w, h in chest_zones],
|
||||
"best_body_centric": best_body,
|
||||
"best_chest_centric": best_chest,
|
||||
"cross_eval": {
|
||||
"body_pl_on_chest": body_pl_on_chest["total_coverage_fraction"],
|
||||
"chest_pl_on_body": chest_pl_on_body["total_coverage_fraction"],
|
||||
"chest_gain_pp": chest_gain_pp,
|
||||
"body_loss_pp": body_loss_pp,
|
||||
},
|
||||
"verdict": verdict,
|
||||
}
|
||||
Path(args.out).parent.mkdir(parents=True, exist_ok=True)
|
||||
Path(args.out).write_text(json.dumps(out, indent=2))
|
||||
print(f"Wrote {args.out}")
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -0,0 +1,118 @@
|
||||
{
|
||||
"room": {
|
||||
"width_m": 5.0,
|
||||
"height_m": 5.0
|
||||
},
|
||||
"freq_ghz": 2.4,
|
||||
"body_zones": [
|
||||
{
|
||||
"name": "bed",
|
||||
"x": 1.5,
|
||||
"y": 0.5,
|
||||
"w": 2.0,
|
||||
"h": 1.5
|
||||
},
|
||||
{
|
||||
"name": "chair",
|
||||
"x": 3.5,
|
||||
"y": 3.5,
|
||||
"w": 0.8,
|
||||
"h": 0.8
|
||||
},
|
||||
{
|
||||
"name": "desk",
|
||||
"x": 0.2,
|
||||
"y": 2.5,
|
||||
"w": 1.0,
|
||||
"h": 0.6
|
||||
}
|
||||
],
|
||||
"chest_zones": [
|
||||
{
|
||||
"name": "bed_chest",
|
||||
"x": 2.2,
|
||||
"y": 0.8,
|
||||
"w": 0.6,
|
||||
"h": 0.4
|
||||
},
|
||||
{
|
||||
"name": "chair_chest",
|
||||
"x": 3.7,
|
||||
"y": 3.7,
|
||||
"w": 0.4,
|
||||
"h": 0.4
|
||||
},
|
||||
{
|
||||
"name": "desk_chest",
|
||||
"x": 0.5,
|
||||
"y": 2.7,
|
||||
"w": 0.4,
|
||||
"h": 0.2
|
||||
}
|
||||
],
|
||||
"best_body_centric": {
|
||||
"score": 0.493006993006993,
|
||||
"tx": [
|
||||
4.25,
|
||||
0.0
|
||||
],
|
||||
"rx": [
|
||||
0.0,
|
||||
3.25
|
||||
],
|
||||
"link_m": 5.350233639758174,
|
||||
"per_zone": {
|
||||
"bed": {
|
||||
"area_m2": 3.0000000000000004,
|
||||
"covered_m2": 1.6175000000000004,
|
||||
"coverage_fraction": 0.5391666666666667
|
||||
},
|
||||
"chair": {
|
||||
"area_m2": 0.6400000000000001,
|
||||
"covered_m2": 0.0,
|
||||
"coverage_fraction": 0.0
|
||||
},
|
||||
"desk": {
|
||||
"area_m2": 0.6500000000000001,
|
||||
"covered_m2": 0.4975000000000001,
|
||||
"coverage_fraction": 0.7653846153846153
|
||||
}
|
||||
}
|
||||
},
|
||||
"best_chest_centric": {
|
||||
"score": 0.8235294117647058,
|
||||
"tx": [
|
||||
2.0,
|
||||
0.0
|
||||
],
|
||||
"rx": [
|
||||
4.5,
|
||||
5.0
|
||||
],
|
||||
"link_m": 5.5901699437494745,
|
||||
"per_zone": {
|
||||
"bed_chest": {
|
||||
"area_m2": 0.29250000000000004,
|
||||
"covered_m2": 0.28750000000000003,
|
||||
"coverage_fraction": 0.9829059829059829
|
||||
},
|
||||
"chair_chest": {
|
||||
"area_m2": 0.20250000000000004,
|
||||
"covered_m2": 0.20250000000000004,
|
||||
"coverage_fraction": 1.0
|
||||
},
|
||||
"desk_chest": {
|
||||
"area_m2": 0.10000000000000002,
|
||||
"covered_m2": 0.0,
|
||||
"coverage_fraction": 0.0
|
||||
}
|
||||
}
|
||||
},
|
||||
"cross_eval": {
|
||||
"body_pl_on_chest": 0.5546218487394958,
|
||||
"chest_pl_on_body": 0.40326340326340326,
|
||||
"chest_gain_pp": 26.890756302521,
|
||||
"body_loss_pp": 8.974358974358976
|
||||
},
|
||||
"verdict": "CHEST-CENTRIC WINS: significant placement-strategy change. R6.2.3 should be a CLI option."
|
||||
}
|
||||
Reference in New Issue
Block a user