Content Understanding 2025-11-01 GA

specification/ai/ContentUnderstanding/audioVisualContent.tsp

@@ -28,7 +28,7 @@ model AudioVisualContent extends MediaContent {
   @doc("Height of each video frame in pixels, if applicable.")
   height?: int32;
 
-  @added(Versions.v2025_05_01_preview)
+  // @added(Versions.v2025_05_01_preview)
   @doc("List of camera shot changes in the video, represented by its timestamp in milliseconds.  Only if returnDetails is true.")
   cameraShotTimesMs?: int64[];
 
@@ -38,17 +38,18 @@ model AudioVisualContent extends MediaContent {
   @doc("List of transcript phrases.  Only if returnDetails is true.")
   transcriptPhrases?: TranscriptPhrase[];
 
-  @removed(Versions.v2025_05_01_preview)
-  @doc("List of faces in the video.  Only if enableFace and returnDetails are true.")
-  faces?: ImageFace[];
+  // @removed(Versions.v2025_05_01_preview)
+  // @doc("List of faces in the video.  Only if enableFace and returnDetails are true.")
+  // faces?: ImageFace[];
 
-  @added(Versions.v2025_05_01_preview)
-  @doc("List of detected persons in the video.  Only if enableFace and returnDetails are true.")
-  persons?: DetectedPerson[];
+  // @added(Versions.v2025_05_01_preview)
+  // @removed(Versions.v2025_11_01)
+  // @doc("List of detected persons in the video.  Only if enableFace and returnDetails are true.")
+  // persons?: DetectedPerson[];
 
-  @added(Versions.v2025_05_01_preview)
-  @doc("List of audio visual segments.  Only if enableSegmentation and returnDetails are true.")
-  segments?: AudioVisualSegment[];
+  @added(Versions.v2025_11_01)
+  @doc("List of detected content segments.  Only if enableSegment is true.")
+  segments?: AudioVisualContentSegment[];
 }
 
 @doc("Transcript phrase.")
@@ -95,34 +96,55 @@ model TranscriptWord {
   span?: ContentSpan;
 }
 
-@added(Versions.v2025_05_01_preview)
-@doc("Audio visual segment, such as a scene, chapter, etc.")
-model AudioVisualSegment {
-  @doc("Segment ID.")
-  segmentId: string;
+// @added(Versions.v2025_05_01_preview)
+// @removed(Versions.v2025_11_01)
+// @doc("Audio visual segment, such as a scene, chapter, etc.")
+// model AudioVisualSegment {
+//   @doc("Segment ID.")
+//   segmentId: string;
 
-  @doc("Start time of the segment in milliseconds.")
-  startTimeMs: int64;
+//   @doc("Start time of the segment in milliseconds.")
+//   startTimeMs: int64;
 
-  @doc("End time of the segment in milliseconds.")
-  endTimeMs: int64;
+//   @doc("End time of the segment in milliseconds.")
+//   endTimeMs: int64;
 
-  @doc("Short description of the segment.")
-  description: string;
+//   @doc("Short description of the segment.")
+//   description: string;
 
-  @doc("Span of the segment in the markdown content.")
-  span?: ContentSpan;
-}
+//   @doc("Span of the segment in the markdown content.")
+//   span?: ContentSpan;
+// }
 
-@added(Versions.v2025_05_01_preview)
-@doc("Detected person.")
-model DetectedPerson {
-  @doc("Person identifier in the optional person directory if found.  Otherwise, each unknown person is assigned a unique `Person-{Number}`.")
-  personId?: string;
+// @added(Versions.v2025_05_01_preview)
+// @removed(Versions.v2025_11_01)
+// @doc("Detected person.")
+// model DetectedPerson {
+//   @doc("Person identifier in the optional person directory if found.  Otherwise, each unknown person is assigned a unique `Person-{Number}`.")
+//   personId?: string;
 
-  @doc("Confidence of the person identification, if a person directory is provided.")
-  confidence?: float32;
+//   @doc("Confidence of the person identification, if a person directory is provided.")
+//   confidence?: float32;
+
+//   @doc("Encoded source that identifies the position of the person in the input content.")
+//   source?: SourceExpression;
+// }
 
-  @doc("Encoded source that identifies the position of the person in the input content.")
-  source?: SourceExpression;
+@added(Versions.v2025_11_01)
+@doc("Detected audio/visual content segment.")
+model AudioVisualContentSegment {
+  @doc("Segment identifier.")
+  segmentId: string;
+
+  @doc("Classified content category.")
+  category: string;
+
+  @doc("Span of the segment in the markdown content.")
+  span: ContentSpan;
+
+  @doc("Start time of the segment in milliseconds.")
+  startTimeMs: int64;
+
+  @doc("End time of the segment in milliseconds.")
+  endTimeMs: int64;
 }