Get predictions from a video action recognition model

Stay organized with collections Save and categorize content based on your preferences.

This page shows you how to get batch predictions from your video action recognition models using the Google Cloud console or the Vertex AI API. Batch predictions are asynchronous requests. You request batch predictions directly from the model resource without needing to deploy the model to an endpoint.

AutoML video models do not support online predictions.

Get batch predictions

To make a batch prediction request, you specify an input source and an output format where Vertex AI stores predictions results.

Input data requirements

The input for batch requests specifies the items to send to your model for prediction. Batch predictions for the AutoML video model type use a JSON Lines file to specify a list of videos to make predictions for, and then store the JSON Lines file in a Cloud Storage bucket. You can specify Infinity for the timeSegmentEnd field to specify the end of the video. The following sample shows a single line in an input JSON Lines file.

{'content': 'gs://sourcebucket/datasets/videos/source_video.mp4', 'mimeType': 'video/mp4', 'timeSegmentStart': '0.0s', 'timeSegmentEnd': '2.366667s'}

Request a batch prediction

For batch prediction requests, you can use the Google Cloud console or the Vertex AI API. Depending on the number of input items that you've submitted, a batch prediction task can take some time to complete.

Google Cloud console

Use the Google Cloud console to request a batch prediction.

  1. In the Google Cloud console, in the Vertex AI section, go to the Batch predictions page.

    Go to the Batch predictions page

  2. Click Create to open the New batch prediction window and complete the following steps:

    1. Enter a name for the batch prediction.
    2. For Model name, select the name of the model to use for this batch prediction.
    3. For Source path, specify the Cloud Storage location where your JSON Lines input file is located.
    4. For the Destination path, specify a Cloud Storage location where the batch prediction results are stored. The Output format is determined by your model's objective. AutoML models for image objectives output JSON Lines files.


Use the Vertex AI API to send batch prediction requests.


Before using any of the request data, make the following replacements:

  • LOCATION: Region where Model is stored and batch prediction job is executed. For example, us-central1.
  • PROJECT: Your project ID
  • BATCH_JOB_NAME: Display name for the batch job
  • MODEL_ID: The ID for the model to use for making predictions
  • THRESHOLD_VALUE (optional): Model returns only predictions that have confidence scores with at least this value
  • URI: Cloud Storage URI where your input JSON Lines file is located.
  • BUCKET: Your Cloud Storage bucket
  • PROJECT_NUMBER: Project number for your project

HTTP method and URL:


Request JSON body:

    "displayName": "BATCH_JOB_NAME",
    "model": "projects/PROJECT/locations/us-central1/MODEL_ID",
    "modelParameters": {
      "confidenceThreshold": THRESHOLD_VALUE,
    "inputConfig": {
        "instancesFormat": "jsonl",
        "gcsSource": {
            "uris": ["URI"],
    "outputConfig": {
        "predictionsFormat": "jsonl",
        "gcsDestination": {
            "outputUriPrefix": "OUTPUT_BUCKET",

To send your request, choose one of these options:


Save the request body in a file called request.json, and execute the following command:

curl -X POST \
-H "Authorization: Bearer $(gcloud auth print-access-token)" \
-H "Content-Type: application/json; charset=utf-8" \
-d @request.json \


Save the request body in a file called request.json, and execute the following command:

$cred = gcloud auth print-access-token
$headers = @{ "Authorization" = "Bearer $cred" }

Invoke-WebRequest `
-Method POST `
-Headers $headers `
-ContentType: "application/json; charset=utf-8" `
-InFile request.json `
-Uri "" | Select-Object -Expand Content

You should receive a JSON response similar to the following:

  "name": "projects/PROJECT_NUMBER/locations/us-central1/batchPredictionJobs/BATCH_JOB_ID",
  "displayName": "BATCH_JOB_NAME 202005291958",
  "model": "projects/PROJECT_NUMBER/locations/us-central1/models/MODEL_ID",
  "inputConfig": {
    "instancesFormat": "jsonl",
    "gcsSource": {
      "uris": [
  "outputConfig": {
    "predictionsFormat": "jsonl",
    "gcsDestination": {
      "outputUriPrefix": "BUCKET"
  "state": "JOB_STATE_PENDING",
  "createTime": "2020-05-30T02:58:44.341643Z",
  "updateTime": "2020-05-30T02:58:44.341643Z",
  "modelDisplayName": "MODEL_NAME",
  "modelObjective": "MODEL_OBJECTIVE"

You can poll for the status of the batch job using the BATCH_JOB_ID until the job state is JOB_STATE_SUCCEEDED.


To learn how to install and use the client library for Vertex AI, see Vertex AI client libraries. For more information, see the Vertex AI Java API reference documentation.


public class CreateBatchPredictionJobVideoActionRecognitionSample {

  public static void main(String[] args) throws IOException {
    // TODO(developer): Replace these variables before running the sample.
    String project = "PROJECT";
    String displayName = "DISPLAY_NAME";
    String model = "MODEL";
    String gcsSourceUri = "GCS_SOURCE_URI";
    String gcsDestinationOutputUriPrefix = "GCS_DESTINATION_OUTPUT_URI_PREFIX";
        project, displayName, model, gcsSourceUri, gcsDestinationOutputUriPrefix);

  static void createBatchPredictionJobVideoActionRecognitionSample(
      String project,
      String displayName,
      String model,
      String gcsSourceUri,
      String gcsDestinationOutputUriPrefix)
      throws IOException {
    JobServiceSettings settings =
    String location = "us-central1";

    // Initialize client that will be used to send requests. This client only needs to be created
    // once, and can be reused for multiple requests. After completing all of your requests, call
    // the "close" method on the client to safely clean up any remaining background resources.
    try (JobServiceClient client = JobServiceClient.create(settings)) {
      Value modelParameters = ValueConverter.EMPTY_VALUE;
      GcsSource gcsSource = GcsSource.newBuilder().addUris(gcsSourceUri).build();
      BatchPredictionJob.InputConfig inputConfig =
      GcsDestination gcsDestination =
      BatchPredictionJob.OutputConfig outputConfig =

      String modelName = ModelName.of(project, location, model).toString();

      BatchPredictionJob batchPredictionJob =
      LocationName parent = LocationName.of(project, location);
      BatchPredictionJob response = client.createBatchPredictionJob(parent, batchPredictionJob);
      System.out.format("response: %s\n", response);
      System.out.format("\tName: %s\n", response.getName());


To learn how to install and use the client library for Vertex AI, see Vertex AI client libraries. For more information, see the Vertex AI Python API reference documentation.

def create_batch_prediction_job_sample(
    project: str,
    location: str,
    model_resource_name: str,
    job_display_name: str,
    gcs_source: Union[str, Sequence[str]],
    gcs_destination: str,
    sync: bool = True,
    aiplatform.init(project=project, location=location)

    my_model = aiplatform.Model(model_resource_name)

    batch_prediction_job = my_model.batch_predict(


    return batch_prediction_job

Retrieve batch prediction results

Vertex AI sends batch prediction output to your specified destination.

When a batch prediction task is complete, the output of the prediction is stored in the Cloud Storage bucket that you specified in your request.

Example batch prediction results

The following an example batch prediction results from a video action recognition model.

  "instance": {
   "content": "gs://bucket/video.mp4",
    "mimeType": "video/mp4",
    "timeSegmentStart": "1s",
    "timeSegmentEnd": "5s"
  "prediction": [{
    "id": "1",
    "displayName": "swing",
    "timeSegmentStart": "1.2s",
    "timeSegmentEnd": "1.2s",
    "confidence": 0.7
  }, {
    "id": "2",
    "displayName": "jump",
    "timeSegmentStart": "3.4s",
    "timeSegmentEnd": "3.4s",
    "confidence": 0.5