Real-time Video Stream Processing for Airport Security | AWS Services for ML

Processing Real-time Video Streams from Airports for Security Event Detection

Prev Question Next Question

Question

You work as a machine learning specialist for the department of defense in the NSA (National Security Agency)

The NSA is responsible for security in the ports of entry around the United States.

You need to process real-time video streams from airports around the country to identify questionable activity within the airport facilities and send the streaming data to SageMaker to be used as training data for your model.

Your model needs to trigger an alert system when a security event is detected.

What AWS services would you use to create this system most accurately and cost-effectively?

Answers

A. Use AWS Rekognition to process your video streams and send the processed data to your SageMaker model. When the model detects a security event, a lambda function is triggered to publish an SNS message to the alert system.

B. Use AWS Elastic Transcoder to process the video streams and send the processed data to your SageMaker model. When the model detects a security event, a lambda function is triggered to publish an SNS message to the alert system.

C. Use Amazon Kinesis Video Streams to stream the video to a set of processing workers running in ECS Fargate. The workers send the video data to your SageMaker machine learning model which identifies alert situations. These alerts are processed by Kinesis Data Streams which uses a lambda function to trigger the alert system.

D. Use Amazon Kinesis Data Streams to process your video data using lambda functions which push out an SNS notification to the alert system when a security event is detected.

Show Answer

Explanations

Click on the arrows to vote for the correct answer

A. B. C. D.

Answer: C.

Option A is incorrect.

The AWS Rekognition service is not meant to process streams.

It works with Kinesis Video Streams to provide this capability.

Also, it needs another component to send its output to your SageMaker model.

This part of the solution is missing.

Option B is incorrect.

The Amazon Elastic Transcoder service is used to convert video files from one format to another.

It would not be useful to stream video to a processing service.

(See the AWS documentation titled Amazon Elastic Transcoder)

Option C is correct.

The Amazon Kinesis Video Streams service will stream your videos to a processing service that feeds your machine learning model running in SageMaker.

Kinesis Streams using lambda to trigger event consumption.(See the AWS machine learning blog titled Analyze live video at scale in real-time using Amazon Kinesis Video Streams and Amazon SageMaker)

Option D is incorrect.

This option lacks the machine learning component of the solution.

Reference:

Please see the Amazon Kinesis Video Streams documentation.

See a depiction of the proposed solution (in the AWS machine Learning blog titled: Analyze live video at scale in real-time using Amazon Kinesis Video Streams and Amazon SageMaker)

The best AWS services to use in this scenario are Amazon Kinesis Video Streams, Amazon SageMaker, and Amazon Kinesis Data Streams, along with AWS Lambda and Amazon SNS.

Option C is the most accurate and cost-effective solution to process real-time video streams from airports around the country to identify questionable activity within the airport facilities and send the streaming data to SageMaker to be used as training data for your model.

Option A is incorrect because Amazon Rekognition is not the best choice for processing the video streams in real-time. Amazon Rekognition is a computer vision service that can identify objects, people, text, and activities in images and videos. It is designed for use cases such as detecting faces and emotions, text recognition, celebrity recognition, and object detection. However, it is not optimized for real-time video processing.

Option B is incorrect because AWS Elastic Transcoder is a media transcoding service that converts media files from one format to another. It is not designed for real-time video processing, and it does not have the capability to identify security events in video streams.

Option D is incorrect because Amazon Kinesis Data Streams is a real-time data streaming service that is used to capture, process, and store data streams in real-time. It can be used with AWS Lambda to process data in real-time and send notifications to the alert system. However, it is not designed for processing video streams, and it does not have the capability to identify security events in video streams.

Amazon Kinesis Video Streams is a real-time video streaming service that makes it easy to ingest and process video streams at scale. It can be used to capture real-time video streams from airports around the country and stream them to a set of processing workers running in Amazon ECS Fargate.

Amazon SageMaker is a fully-managed service that provides developers and data scientists with the ability to build, train, and deploy machine learning models at scale. It can be used to build a machine learning model that can identify security events in real-time video streams.

Amazon Kinesis Data Streams can be used to process alerts generated by the machine learning model in real-time. AWS Lambda can be used to process the alerts and send notifications to the alert system using Amazon SNS.

In summary, option C is the most accurate and cost-effective solution for processing real-time video streams from airports around the country to identify security events and send notifications to the alert system. The solution combines Amazon Kinesis Video Streams, Amazon SageMaker, Amazon Kinesis Data Streams, AWS Lambda, and Amazon SNS to create an end-to-end video processing and alerting system.

Prev Question Next Question