Does sagemaker support reading video input from s3 for BatchTransform jobs?

[velociraptor111]

Thanks again for creating this great product. I have a couple of questions:

1. Can we read in video data as input to our batch transform jobs?
2. What is the limit to the processing time per-input for batch transform jobs? How can we increase this?
3. If 1) is currently not supported, how do you suggest we achieve this functionality? Is it a good idea to download the video from S3 ourselves and go on from there?

Related Question:
aws/sagemaker-mxnet-inference-toolkit#77

Thanks

[nadiaya]

Thanks for using SageMaker! I've reached out to the relevant team.

[amitpeshwani]

Hi, @nadiaya any thoughts on using Batch transform for large video inference?