Fix Meet visual: reject solid avatar tiles + screen-share OCR

Root cause of the "4 people → 2 speakers" Meet call: the colored-border detector
read solid camera-off avatar tiles (orange "J", magenta "G") as active speakers
for the ENTIRE call. Those whole-call phantom spans dominated backend name
attribution, collapsing every remote voice onto one name — and the giant filled
bbox also swallowed screen-share text (WERUNBTC.COM ×49) as a speaker.

Validated against 9 real fixtures (harness over the real MeetAdapter):

Detection:
- FrameSampler.thinColoredPoints: coloured counterpart of thinWhitePoints — keeps
  thin border/ring/pill edges, drops solid colour fills.
- GridCallAnalyzer.isHollow: reject a highlight component whose interior is filled
  (a solid tile) vs a hollow ring (a real border). Config.maxInteriorFill (0.2 default).
- MeetAdapter: detect thin BLUE edges only (hue 180–240°, measured from the
  fixtures), maxInteriorFill 0.3 (real Meet rings ≈0.2–0.3, solid tiles ≈0.36).
- Result on fixtures: John Arnold/Grant Gilliam (solid tiles) now NEVER detected;
  Matt Odell/Mark detected when their blue cue is present. Sparse but never wrong —
  correct for a naming hint over audio diarization.

OCR name hygiene:
- isLikelyName rejects domain-like screen-share text ("WERUNBTC.COM", OCR'd ".GOM").
- cleaned() strips trailing punctuation ("Mark." → "Mark").
- TimelineBuilder.canonicalizeByFrequency folds rare OCR misspellings into a
  dominant near-twin name ("Matt Odel"/"MattOdell" → "Matt Odell", "Mare" → "Mark").

Tests: hollow-ring, extended OCR filter, fuzzy-merge. 65 pass.
This commit is contained in:
Grant Gilliam
2026-06-08 16:18:52 -05:00
parent 5c80e827a1
commit 39beccf7f4
6 changed files with 182 additions and 6 deletions
@@ -32,6 +32,16 @@ struct MeetAdapter: AppAdapter {
// The bright ring (#1a73e8) is ~0.89 sat but the lighter glow (#8ab4f8) is
// ~0.44, below the 0.5 default lower the threshold so the glow registers.
config.colorSaturation = 0.35
// Meet's active cue is a thin BLUE (210°) ring + audio pill. Detect thin blue
// EDGES only, gated to blue: this rejects solid camera-off avatar tiles (orange
// 30°, magenta 340°), which otherwise read as "speaking" for the whole call
// and collapse every remote voice onto one name. Validated on real fixtures.
config.coloredBorderThinOnly = true
config.colorHueRange = 180...240
// Meet's blue border is faint; real rings measure 0.200.30 interior fill while
// solid tiles measure 0.36, so allow a higher fill here than the 0.2 default to
// recover real borders without readmitting the solid-tile false positives.
config.maxInteriorFill = 0.3
config.tileExpandX = 3.0
config.tileExpandY = 5.0
self.analyzer = GridCallAnalyzer(config: config)