Cloud Video Intelligence rolled out improved models for video annotation using the OBJECT_TRACKING and
Share
Services
## Feature
Cloud Video Intelligence rolled out improved models for video annotation using the [OBJECT\_TRACKING](https://cloud.google.com/video-intelligence/docs/object-tracking) and[TEXT\_DETECTION](https://cloud.google.com/video-intelligence/docs/text-detection) with the v1 version of the service.
## Feature
You can now specify which model you want to use with [LABEL\_DETECTION](https://cloud.google.com/video-intelligence/docs/analyze-labels) and[SHOT\_CHANGE](https://cloud.google.com/video-intelligence/docs/analyze-shots). To specify a model using the v1 version of the service, set the `model` field of the[LabelDetectionConfig](https://cloud.google.com/video-intelligence/docs/reference/rest/v1/videos/annotate#labeldetectionconfig)or[ShotChangeDetectionConfig](https://cloud.google.com/video-intelligence/docs/reference/rest/v1/videos/annotate#shotchangedetectionconfig)to either `builtin/stable` or `builtin/latest`.
## Feature
You can now specify your own custom AutoML model when performing[label detection](https://cloud.google.com/video-intelligence/docs/streaming/label-analysis) or[object tracking](https://cloud.google.com/video-intelligence/docs/streaming/object-tracking)on a stream. This feature is in beta.