Amazon Kinesis Video Streams

We have to process the video stream to identify emotions. For which we need to use Amazon Rekognition/custom model on Amazon SageMaker. With Kinesis WebRTC Javascript SDK, currently, video can be streamed only into the kinesis signaling channel. Signaling channel data is available for streaming only and not processing (ML). So, how can we get real-time data for processing into Kinesis Streams from the frontend?

For streaming the video from frontend to backend into the Amazon Kinesis Video Streams for processing, we tested with Kinesis webRTC JavaScript SDK, and we are facing issues while implementing as mentioned above, so would Chime SDK serve as an alternative to this?

In Rekognition, "create-stream-processor" has a settings parameter. This currently only supports FaceSearch. We are looking to Detect and analyze faces. Is that possible with "create-stream-processor" in the Python SDK? Or do we have to use the Java SDK?

We would like to connect a number of (about 25) video streams, from an Amazon S3 bucket containing video data to endpoints accessible to a Docker image, which, when run, will process the input video streams and emit some JSON statistics.

The 25 video streams should be synchronized. Could people share their experiences with a similar scenario and perhaps offer advice about which is better (Wowza, Amazon Kinesis Video Streams) for this kind of problem, or why they chose one technology over the other?

The video stream duration will be quite long (about 8 hours each x 25 camera sources). The 25 video streams will have no audio component. If you worked with a similar problem, what was your experience with scaling, latency, resource requirements, config, etc.?