diff --git a/packages/google-cloud-videointelligence/protos/google/cloud/videointelligence/v1p2beta1/video_intelligence.proto b/packages/google-cloud-videointelligence/protos/google/cloud/videointelligence/v1p2beta1/video_intelligence.proto
index 2c1af2efddd..ce540c90c08 100644
--- a/packages/google-cloud-videointelligence/protos/google/cloud/videointelligence/v1p2beta1/video_intelligence.proto
+++ b/packages/google-cloud-videointelligence/protos/google/cloud/videointelligence/v1p2beta1/video_intelligence.proto
@@ -352,6 +352,13 @@ message NormalizedBoundingPoly {
 message TextSegment {
   // Video segment where a text snippet was detected.
   VideoSegment segment = 1;
+
+  // Confidence for the track of detected text. It is calculated as the highest
+  // over all frames where OCR detected text appears.
+  float confidence = 2;
+
+  // Information related to the frames where OCR detected text appears.
+  repeated TextFrame frames = 3;
 }
 
 // Video frame level annotation results for text annotation (OCR).
@@ -372,15 +379,8 @@ message TextAnnotation {
   // The detected text.
   string text = 1;
 
-  // Confidence for the track of detected text. It is calculated as the highest
-  // over all frames where OCR detected text appears.
-  float confidence = 2;
-
-  // Information related to the frames where OCR detected text appears.
-  repeated TextFrame frames = 3;
-
   // All video segments where OCR detected text appears.
-  repeated TextSegment segments = 4;
+  repeated TextSegment segments = 2;
 }
 
 // Video frame level annotations for object detection and tracking. This field
diff --git a/packages/google-cloud-videointelligence/src/v1p2beta1/doc/google/cloud/videointelligence/v1p2beta1/doc_video_intelligence.js b/packages/google-cloud-videointelligence/src/v1p2beta1/doc/google/cloud/videointelligence/v1p2beta1/doc_video_intelligence.js
index 017dba339cf..93622c9836d 100644
--- a/packages/google-cloud-videointelligence/src/v1p2beta1/doc/google/cloud/videointelligence/v1p2beta1/doc_video_intelligence.js
+++ b/packages/google-cloud-videointelligence/src/v1p2beta1/doc/google/cloud/videointelligence/v1p2beta1/doc_video_intelligence.js
@@ -550,6 +550,15 @@ const NormalizedBoundingPoly = {
  *
  *   This object should have the same structure as [VideoSegment]{@link google.cloud.videointelligence.v1p2beta1.VideoSegment}
  *
+ * @property {number} confidence
+ *   Confidence for the track of detected text. It is calculated as the highest
+ *   over all frames where OCR detected text appears.
+ *
+ * @property {Object[]} frames
+ *   Information related to the frames where OCR detected text appears.
+ *
+ *   This object should have the same structure as [TextFrame]{@link google.cloud.videointelligence.v1p2beta1.TextFrame}
+ *
  * @typedef TextSegment
  * @memberof google.cloud.videointelligence.v1p2beta1
  * @see [google.cloud.videointelligence.v1p2beta1.TextSegment definition in proto format]{@link https://github.com/googleapis/googleapis/blob/master/google/cloud/videointelligence/v1p2beta1/video_intelligence.proto}
@@ -589,15 +598,6 @@ const TextFrame = {
  * @property {string} text
  *   The detected text.
  *
- * @property {number} confidence
- *   Confidence for the track of detected text. It is calculated as the highest
- *   over all frames where OCR detected text appears.
- *
- * @property {Object[]} frames
- *   Information related to the frames where OCR detected text appears.
- *
- *   This object should have the same structure as [TextFrame]{@link google.cloud.videointelligence.v1p2beta1.TextFrame}
- *
  * @property {Object[]} segments
  *   All video segments where OCR detected text appears.
  *