Class SpeakerSegment
- Namespace
- VisioForge.Core.Types.X.AI
- Assembly
- VisioForge.Core.dll
One diarized speech turn: a time span attributed to a single speaker.
public class SpeakerSegmentInheritance
Inherited Members
Constructors
SpeakerSegment(int, TimeSpan, TimeSpan, double)
Initializes a new instance of the VisioForge.Core.Types.X.AI.SpeakerSegment class.
public SpeakerSegment(int speakerId, TimeSpan start, TimeSpan end, double confidence)Parameters
speakerIdint-
The stable, zero-based speaker identifier.
startTimeSpan-
The start time of the turn on the media timeline.
endTimeSpan-
The end time of the turn on the media timeline.
confidencedouble-
The turn's agreement score: the mean fraction of covering segmentation windows that voted for the assigned speaker (0..1).
Properties
Confidence
Gets the turn's agreement score in [0, 1]: the mean, over the turn's frames, of the fraction of covering segmentation windows that voted for the assigned speaker. It is a measured agreement, not a model score — 1.0 means every window that overlapped the moment named the same speaker.
public double Confidence { get; }Property Value
Duration
Gets the duration of the turn.
public TimeSpan Duration { get; }Property Value
End
Gets the end time of the turn on the media timeline.
public TimeSpan End { get; }Property Value
SpeakerId
Gets the stable, zero-based speaker identifier this turn was attributed to.
public int SpeakerId { get; }Property Value
Start
Gets the start time of the turn on the media timeline.
public TimeSpan Start { get; }Property Value
Methods
ToString()
Returns a readable representation of the turn.
public override string ToString()Returns
- string
-
A string describing the speaker id and time span.