Hey @OlafenwaMoses and @john,
Does Deepstack currently have the ability to process/analyze video (either streaming video or file-based) in addition to single-frame images?
Example clips from other software demonstrating objects being detected, identified, and outlined throughout every frame of the clip make me want to do the same with Deepstack.
Examples:
Would love to hear your thoughts on this capability now and in the future. Thanks! 
Yes it can. In the end video are made in frames per second. If you are using CPU it may take 250ms to analyze one frame, so you could analyze video at 4 fps. With GPU you would be able to analyze 30 fps.
I’m pretty sure I saw a code about using deepstack to analyze video from a webcam but wasn’t able to find it right now.
There are a number of third party software tools that will use deepstack to process frames, but deepstack itself does not handle video. since deepstack is a rest API I dont see it every handling video directly
Makes sense, Robin. That was my assumption.
@aesterling @robmarkcole @mrpie Thanks for creating this discussion.
This is an issue we have started working on to address to reduce the steps involved in using/deploying DeepStack.
In a couple of days, @john will publish and share details of the DeepStack Python SDK he just developed ( more languages to follow) which lets you easily do the following in your Python code,
Excellent, that sounds great, and can’t wait to hear more! Please follow up here when you have more to share.
Hey @OlafenwaMoses and @john, hope you are both well. I’m checking back in on this for an update, please, specifically about processing a video source. Thanks!
@aesterling the library is out and work ongoing to add sufficient documentation.
For video processing, you can check the sample code inthis file ( samples will be updated soon too )
@aesterling Do check this out.
DeepStack Python SDK now available - Announcement - DeepStack Forum