FirebaseAILogic Framework Reference

ActivityDetectionConfig

public struct ActivityDetectionConfig : Sendable

Configures the model's automatic detection of user activity.

Public Preview: This API is a public preview and may be subject to change.

Seealso

RealtimeInputConfig

• Sensitivity

  How sensitive the model interprets speech activity.

  Public Preview: This API is a public preview and may be subject to change.

  Declaration

  Swift

  public enum Sensitivity : Sendable

• init(startSensitivity:endSensitivity:prefixPadding:silenceDuration:)

  Creates a new ActivityDetectionConfig value.

  Declaration

  Swift

  public init(startSensitivity: Sensitivity? = nil,
              endSensitivity: Sensitivity? = nil,
              prefixPadding: TimeInterval? = nil,
              silenceDuration: TimeInterval? = nil)

  Parameters

  startSensitivity	Determines how likely the start of speech is detected.
  endSensitivity	Determines how likely the end of speech is detected.
  prefixPadding	How long detected speech should be present before start-of-speech is committed.
  silenceDuration	How long silence (or non-speech) should be present before end-of-speech is committed.

• init(startSensitivity:endSensitivity:prefixPadding:silenceDuration:)

  Creates a new ActivityDetectionConfig value.

  This method uses Duration for timing related properties instead of TimeInterval. See the "See Also" section for the alternative initializer if you don't have access to the Duration API.

  Declaration

  Swift

  @available(iOS 16.0, macOS 13.0, tvOS 16.0, watchOS 9.0, *)
  public init(startSensitivity: Sensitivity? = nil,
              endSensitivity: Sensitivity? = nil,
              prefixPadding: Duration? = nil,
              silenceDuration: Duration? = nil)

  Parameters

  startSensitivity	Determines how likely the start of speech is detected.
  endSensitivity	Determines how likely the end of speech is detected.
  prefixPadding	How long detected speech should be present before start-of-speech is committed.
  silenceDuration	How long silence (or non-speech) should be present before end-of-speech is committed.

• disabled()

  Disables automatic activity detection.

  When automatic activity detection is enabled, the model will interpret detected voices and text as the start of activity.

  When automatic activity detection is disabled, the user must send activity signals manually.

  Seealso

  sendStartActivityRealtime()

  Seealso

  sendStopActivityRealtime()

  Declaration

  Swift

  public static func disabled() -> ActivityDetectionConfig