Adapters: add Meet, Zoom, Teams (coloured border) + adapter registry

Front-loads the remaining visual adapters per the Signal→Meet→Zoom priority.
All three reuse GridCallAnalyzer's coloured-border (saturated) detection path
and share the new bottom-left name anchor:

- GridCallAnalyzer: generalise nameAtBottom:Bool into a NameAnchor enum
  (.bottomCenter for Signal's centered footer, .bottomLeft for Meet/Zoom/Teams
  where the name hugs the tile's bottom-left corner, .center for completeness).
  tileRect estimates the tile up-and-right of a bottom-left name.
- MeetAdapter (Google-blue ring, browser-hosted), ZoomAdapter (green/yellow
  border), TeamsAdapter (violet ring): coloured-border on, white-border off,
  bottom-left names. Geometry constants are first-pass pending real fixtures.
- AdapterRegistry.adapter(for:) maps CallDetector.DetectedApp → AppAdapter so
  VisualObserver can be constructed when live visual capture is wired in;
  unmapped apps degrade to audio-only.

Synthetic 4-tile tests: Meet picks each blue-bordered speaker with no
adjacent-tile bleed, Zoom picks the green-bordered speaker, and Signal's
white-only detector correctly ignores a coloured border. 18/18 XCTest pass.
This commit is contained in:
Grant Gilliam
2026-06-06 09:57:53 -05:00
parent a56b47143c
commit c347acbd97
7 changed files with 230 additions and 9 deletions
+23 -8
View File
@@ -13,10 +13,18 @@ import CoreImage
/// Geometry (`Config`) is a first pass; tile expansion calibrates per app against
/// real screenshot fixtures. Detection *logic* is validated on synthetic frames.
struct GridCallAnalyzer {
/// Where the name label sits relative to its participant tile drives how the
/// tile rect is estimated from the OCR'd name box.
enum NameAnchor {
case bottomCenter // Signal: centered footer; tile extends UP, centered on the name
case bottomLeft // Meet/Zoom: name in the bottom-left corner; tile extends UP and RIGHT
case center // name centered inside the tile
}
struct Config {
var tileExpandX = 2.4 // tile width name width × this
var tileExpandY = 4.8 // tile height name height × this
var nameAtBottom = true // Signal/most: name footer sits at the tile bottom
var nameAnchor: NameAnchor = .bottomCenter
var detectColoredBorder = true
var detectWhiteBorder = true
var minTextConfidence: Float = 0.3
@@ -91,20 +99,27 @@ struct GridCallAnalyzer {
return CGRect(x: box.minX * W, y: (1 - box.maxY) * H, width: box.width * W, height: box.height * H)
}
/// Estimate the participant tile from the name label. With `nameAtBottom`, the
/// tile extends UP from the footer (Signal); otherwise it's centred on the name.
/// Estimate the participant tile from the name label, per the app's `nameAnchor`:
/// - `.bottomCenter` (Signal): tile extends UP from a centered footer.
/// - `.bottomLeft` (Meet/Zoom): name hugs the tile's bottom-left corner; the
/// tile extends UP and to the RIGHT of it.
/// - `.center`: tile centered on the name.
private func tileRect(_ box: CGRect, _ w: Int, _ h: Int) -> CGRect {
let W = Double(w), H = Double(h)
let name = pixelRect(box, w, h)
let nw = name.width * config.tileExpandX
let nh = name.height * config.tileExpandY
let cx = name.midX
let rect: CGRect
if config.nameAtBottom {
switch config.nameAnchor {
case .bottomCenter:
let bottom = name.maxY + name.height * 0.3
rect = CGRect(x: cx - nw / 2, y: bottom - nh, width: nw, height: nh)
} else {
rect = CGRect(x: cx - nw / 2, y: name.midY - nh / 2, width: nw, height: nh)
rect = CGRect(x: name.midX - nw / 2, y: bottom - nh, width: nw, height: nh)
case .bottomLeft:
let bottom = name.maxY + name.height * 0.3
let left = name.minX - name.height * 0.4 // small left padding the corner gutter
rect = CGRect(x: left, y: bottom - nh, width: nw, height: nh)
case .center:
rect = CGRect(x: name.midX - nw / 2, y: name.midY - nh / 2, width: nw, height: nh)
}
return rect.intersection(CGRect(x: 0, y: 0, width: W, height: H))
}