Custom Wake Word Part 1: Capturing Data | TinyML

Поделиться
HTML-код
  • Опубликовано: 22 ноя 2024

Комментарии • 21

  • @alakazellae-commerce9627
    @alakazellae-commerce9627 3 года назад

    I wrote my comment before watching the video. OMG, the part at 3:08 "When the cartoons and video games when the characters would power up and shout the name of their move that they're performing because shows how that makes the move work...?" The joy in Shawn's face of him explaining the topic. Just awsome!

  • @psd4942
    @psd4942 4 года назад +2

    Respected *Shawn Hymel* sir,
    I am a student, and literally the way you have teached electricity and electromagnetism is infinite wonderful!❤️ You have given us knowledge in the best way which nobody has ever given to us!
    In india , teachers don't teach us correctly, they just make. Us gulp the text written in the book.
    I request you teacher Shawn hymel that please try to teach us more physics like that you did in past(electricity, electromagnetism)
    I know making more such videos will take a lot of effort! But please try to help,
    Hope you will reply sir🙏🙏🙏👍❤️😁

  • @alakazellae-commerce9627
    @alakazellae-commerce9627 3 года назад

    Shawn isn't only handsome, he is also very smart. : ) I enjoyed our videos on the SparkFun youtube channel.

  • @delezha380
    @delezha380 4 года назад

    the Edge AI Anomaly Detection series is cool. I'm looking forward to your new project with TinyML. Thanks for your video since that's exactly what I want to see while resting after a whole busy day.

    • @ShawnHymel
      @ShawnHymel  4 года назад

      Glad to hear you like them! Thank you 😊

  • @userou-ig1ze
    @userou-ig1ze 4 года назад

    Come on internetz, why the * does this video only have 33 likes. That's insane! You deserve more! Keep on making and teaching!

  • @amirmahdisoltani1
    @amirmahdisoltani1 4 года назад

    Hi Shawn, your TinyML video series is great, and I really enjoyed watching it during quarantine😷🤒.
    Thanks for good tutorials you share with us.🙏🤩
    To help you, I shared botwords with all of my friends. I hope this has been helpful to you.
    Good luck.✌️

    • @ShawnHymel
      @ShawnHymel  4 года назад

      Awesome! And thank you for sharing botwords. I'm hoping we get a good response :)

  • @omaryanas435949
    @omaryanas435949 4 года назад

    Love the video, Machine learning for fun!, thats awesome

  • @doubleHLabs
    @doubleHLabs 4 года назад +1

    Very cool, I added mine

  • @amrul3076
    @amrul3076 4 года назад +1

    Keep making new video please..

  • @rafa_dzto
    @rafa_dzto 4 года назад

    TinyML on Arduino? omg, amazing!
    I was thinking of training a simple neural network in a Raspberry Pi with a bunch of generated data from different Arduinos, and then transmit the data of the trained net to those arduinos. What do you think?

    • @ShawnHymel
      @ShawnHymel  4 года назад +1

      I wouldn't recommend doing the training on the Pi. While it's possible, it'll be really slow. If you're familiar with any of the deep learning frameworks, your best bet is to use your PC or something like Google Colab (for Tensor flow and/or Keras). Once you have a trained model, you can use it to perform inference on the Pi (easier, faster) or an Arduino (harder, slower, but less power usage). Hope that helps!

    • @rafa_dzto
      @rafa_dzto 4 года назад

      ​@@ShawnHymel Ty for the advice! I'll consider a more powerful PC/server instead :)

  • @Asyss_Complex
    @Asyss_Complex 4 года назад

    Interesting! By the way, what is that book on you desk? I've seen that cover before, a Python course?

    • @ShawnHymel
      @ShawnHymel  4 года назад

      "Hands-On Machine Learning" by Aurélien Géron. I'm about 1/3rd of the way through it--highly recommended, but not as an intro to ML :)

  • @BeefIngot
    @BeefIngot 3 года назад

    Oi, the site is still up. Is the project dead or expanding?

    • @ShawnHymel
      @ShawnHymel  3 года назад +1

      Thanks for the heads up--I took the page down. I've gotten keyword spotting to work through Edge Impulse, but the captured samples did come in handy.