Hi, this is Julien from Arcee. In this video, I want to show you a quick demo of streaming a live video from a Raspberry Pi into Amazon Rekognition Video, and we'll see if we can recognize my face in the stream. So I've got a Raspberry Pi here with a Pi camera, and I built a Kinesis client that is going to capture the video and then stream it into Kinesis Video Streams. All instructions for this actually come from a superb blog post by one of my colleagues, and I will put the URL into the video description.
So let's start this client. I'm going to stream at 320 by 200 resolution, 256 kilobit bitrate. Here we go. Okay, so we can see Kinesis actually receiving the video fragments here, being pushed by the Kinesis client. In a few seconds, when those things get to AWS, as you can see on my screen over there, let's give it a few seconds to stabilize. You can see my live video in the Kinesis Video Streams console, right? It says 320 by 200, the bitrate is a little over 270 kilobits, and we have about 33 seconds delay, which is the amount of time it takes to send all this stuff into Kinesis and for the console to process it. There's just a little bit of delay, but again, I don't have a lot of bandwidth right here. You might hear some beeping noises; it's actually emails being sent through SNS because Rekognition Video is detecting faces in this stream. So, let me look at the camera and see if we can actually see some of those emails saying I'm in here, right? Okay, this should be enough. Yes, and you have to trust me. It says one known person, right? And this is me because I built a face collection with my face in there. So Rekognition Video actually matches my face from the video into the collection.
So there you go, really short video. If you want to replicate this, the only thing you need is a Raspberry Pi, a camera, read the blog post that is referenced in the description, and within maybe an hour or two, you can get this thing running too. It's really cool. You can build lots of cool things with this. That's it for today. Thank you. Bye bye.