setActiveModels method
Enable/disable individual models' per-frame inference at runtime. All
predictors stay loaded in memory, so toggling is instant with no reload —
use it to cut inference load by running only the models a given phase
needs. detect = primary detect, classify = classify model,
secondDetect = the second detect model. Pass null to leave unchanged.
Implementation
Future<void> setActiveModels({
bool? detect,
bool? classify,
bool? secondDetect,
}) async {
final ch = _channel;
if (ch == null) return;
await ch.invokeMethod<void>('setActiveModels', {
if (detect != null) 'detect': detect,
if (classify != null) 'classify': classify,
if (secondDetect != null) 'secondDetect': secondDetect,
});
}