Intel® Movidius™ Neural Compute Stick | IoT Developer Show | Intel Software



hello I'm Martin Kronberg and this is the IOT developer show where we look at IOT technology share learning opportunities and showcase cool demos and the creators behind them in this episode we talk with stuart christie an outtie evangelist here at intel as he walks us through two demos showcasing the movie dias neural computing Stuart thank you so much for joining us today do you want to tell us a little bit more about yourself sure I'm an IT evangelist like yourself Marvin and I'm the hardware guy in this software group on a personal side I'm interested in robotics and photography this is a sort of photography example here we're going to be showing fantastic well I'm really interested in this demo you want to tell us a bit more about it sure it's a follow on from last month's example where we were doing the unboxing of the up square board this time we were showing some demos that you can run on this board using these madea's compute sticks these are also development devices from Intel much smaller as you can see here the USB devices there's an SDK available if you want to write your own programs but today we're just going to demo some of the prepackaged applications that you can find on github and then Meridius application resume sure so kind of like monkeys yes zoom as in animals in the zoo that's correct in fact they call it a park-like area in which deep learning models are housed for exhibition I'm not sure why they call it the zoo but a quick search and github for zoo and Python shows a lot of deep learning examples excellent so what kind of applications are housed in the zoo then we don't have any monkeys but we do have some of birds but today we're going to demo two of them will feature you and me in the zoo I'm using a USB camera to do object recognition and then the second example shows the scalability of these USB sticks and we're on the same demo side by side one running on one stick and one owning in on the second so you can see the speed difference and so all these demos are running on pre frame models exactly the first one using the webcam has been trained on a selection of objects it's fun to get it confused there are Windsor ties lab coats and other seemingly random objects that show up this demo doesn't do any background removal processes the whole video frame so we'll likely see some more guests one of my favorites I've been trying to get is to recognize my face and then cover it up it looks like a ski mask just it thinks it's asking us well let's see what it recognizes here right now I've got a bulletproof vest on it sees me as a person that's good I'm back as a person that yeah interesting so I'm assuming that you can also tweak the threshold levels to kind of get it to be more exactly exactly this one's actually running a threaded example so it's running tiny you're on one of these threads and then one of these USB sticks and then the second one is running Google man and you can tweak the thresholds for it for both of those things using the keyboard here okay and Yolo is you only look once and then it's being used for object detection and then the Google net is being used as an object classification essentially exactly it refines the models yes there are 20 things that the element can detect okay well do you want to take a look at the next demo yes we'll stop the close this one down and then we'll start it up and so you can see with these two stings there's a one stick so this is the Apparel compute demo and you're saying that it's using three of these NVIDIA sticks exactly two sticks here versus one stick and it shows how the system is scalable this demo process is a system of directory of images there's about 65 images and it uses scans an image users OpenCV to print the object name that it's finding on the image and then it also run the stairs running one stick running into sticks and you can see this one the framerate is a lot faster the scrolling years it detects them it's a stick one stick to their yaps take one stick to and then this one's running on stick zero so this shows the scalability you can add as many stakes as you need okay and can I add any photos to this database or is it only trained to recognize these 65 and no you can add as many photographs as you want we run on my own fine and what obviously they've trained it to recognize certain places so know if I've got a cathedral photograph and they haven't trained it to find a Cathedral it's not going to find it it might say that it's a porcupine er so if you find it or a dog yes because it does do probabilities this is what it thinks it is so how is it actually doing the parallel processing its simply doing a ping-pong alternating between the sticks as all three sticks are loaded with the same detection algorithms I've seen other demos where the controlling program pulls the stakes and looks for a free sticks to use that's a better algorithm to use for example if the time to recognize something is indeterminate from see example okay is there a limit to how many sticks you can use a parallel I'm not really a USB port USB 3 port can support for I believe that's the bandwidth limit on the port the photo we've got a photograph we can show you on the screen right now it's got eight it's a bit of a problem to move there's an external PCI to USB card running on an extender and so no it's in windy but you can't use it okay so you mentioned that this is a development platform now are there any kind of production versions of this thing out there there are it's shipping in the Google clip camera and the DJI drone so there's the chip itself there's also a hardware developer kit which is a bit more no look like a regular PCB for doing software development oh I think this is a great platform for developers I want to get into computer vision and thank you so much for coming down here and show it off to us well I've enjoyed showing it off thanks very much for inviting me in February until technologies will help fans experience the Olympic Winter Games Pyeongchang 2018 with the latest innovations in virtual reality 5g Thrones and gaming check out the links to learn about Intel at the Olympic Games watching the algae developer show don't forget to Like and subscribe and remember to check out the links provided to learn more thanks guys

7 Comments

  1. not sure said:

    starts at 0:45

    June 25, 2019
    Reply
  2. Ludwig Schreier said:

    "tweak" to get it "more exact" :-)))))))))))))))))

    June 25, 2019
    Reply
  3. TransistorSnack said:

    0:50 "where we were doing the unboxing of the of [sic] squidward"

    June 25, 2019
    Reply
  4. Andres said:

    2:26 what da little mistake hahaha: a person detected like a bulletproof!! (that robot will shoot you)

    June 25, 2019
    Reply
  5. oh my gosh ponies said:

    Can you do an update for the 2nd generation stick

    June 25, 2019
    Reply
  6. Agata Fors said:

    Для чего нужен этот предмет??? Нужно больше данных!!!

    June 25, 2019
    Reply
  7. Rick Ooi said:

    Hi, is there any limitation in image size use for training, like 25Megapixel? and is it posible for Movidius to process large image size. thanks.

    June 25, 2019
    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *