Recording super slow motion video at high resolution has always been expensive, requiring specialist equipment and the results can be of low, jittery quality at more than 240fps. Using machine learning models like Google Research’s Frame Interpolation for Large Motion (FILM and the Amazon SageMaker cloud service, a new system from AWS “analyses motions between input frames to generate new transitional frames that create seamless ultra-high frame rate slow motion” at low latency (near real-time).
They show how to …
- use FILM and Amazon SageMaker and package the model
- deploy the model at scale using Amazon SageMaker Asynchronous endpoint
- invoke the endpoint to process the video frames
- assemble the original and synthesized frames into a high frame rate slow motion video
Visit the AWS Blog to find out more.