StartLabelDetection

Source: src/AWS/Rekognition/StartLabelDetection.ts

Runtime binding for rekognition:StartLabelDetection — start asynchronous detection of labels (objects, scenes, activities) in a stored video.

Rekognition Video jobs read the video from S3 and run asynchronously — poll the paired Get* binding with the returned JobId, or pass a NotificationChannel (SNS topic + IAM role the caller must be allowed to pass) to be notified on completion.

The binding takes no arguments and grants the function rekognition:StartLabelDetection on * (Rekognition data-plane resources such as collections, users, and jobs are routinely created at runtime, so their identifiers are unknown at deploy time). Provide the implementation with Effect.provide(AWS.Rekognition.StartLabelDetectionHttp).

Video Analysis

// init
const startLabelDetection = yield* AWS.Rekognition.StartLabelDetection();

// runtime
const started = yield* startLabelDetection({
  Video: { S3Object: { Bucket: "videos", Name: "input.mp4" } },
});
// started.JobId