OpenAI Whisper C++ Raspberry Pi 5 Voice Assistant

Поделиться
HTML-код
  • Опубликовано: 18 янв 2025

Комментарии • 15

  • @justward354
    @justward354 11 месяцев назад +2

    Thank you so much! For the past 2 weeks me and my group was stuck not knowing what’s wrong and I started to be afraid to repeat the course. All for that new gpiochip I couldn’t connect any of my sensors.

  • @slow-1978
    @slow-1978 Год назад +4

    This is so freakin cool. You the man. I also own Pi 4 and 5.

  • @yuwaikoon
    @yuwaikoon Год назад +1

    thanks for your great video, any way to run other customized script?

    • @samwechsler
      @samwechsler  Год назад

      Probably. Do mean run a customized script when the spoken phrase is detected? Let me know what you are looking to do.

  • @Tyrone-Ward
    @Tyrone-Ward Год назад +1

    Can you user a screen recorder? It's so hard to see what you're doing sometimes.

    • @samwechsler
      @samwechsler  Год назад +1

      It's been a while since I used a screen recorder. The audio mixing can be annoying. It's nice going back and forth between showing the Pi connections and the screen. Might have to try new camera settings. I understand your view.

    • @Tyrone-Ward
      @Tyrone-Ward Год назад

      @@samwechsler that's fair. In any case, thanks for making these videos. They are very helpful.

  • @Hazar-bt6nf
    @Hazar-bt6nf 6 месяцев назад

    Can it be implemented by Python? Can raspberry pi5 run whisper?

  • @kamertonaudiophileplayer847
    @kamertonaudiophileplayer847 Год назад

    It's better to use the green wire for the green light and the red wire for the red light and finally the yellow one for the ground.

  • @kamertonaudiophileplayer847
    @kamertonaudiophileplayer847 Год назад

    I assume all phrases can be in French.

    • @samwechsler
      @samwechsler  Год назад +1

      If the standard models are run then the French phrases will be translated to English.

  • @ronaldhofman1726
    @ronaldhofman1726 Год назад +1

    my soundblaster package from the 80 , could do speech regocnition, this is a lame example for using AI i think

    • @polerix
      @polerix Год назад

      We had a parrot, desktop commands and Dr. SBAITSO en.wikipedia.org/wiki/Dr._Sbaitso

    • @brulsmurf
      @brulsmurf Год назад +1

      Yeah, but it sucked. If whisper can achieve high accuracy while dealing with accents, background noise etc. Then it actually might be useful.

  • @ArnaudMEURET
    @ArnaudMEURET Год назад +2

    The whole point of the LLM approach to speech recognition is to free us from the. CHOPPY. KEYWORD. MODALITY. that we’ve already mastered on common PCs for 25 years. The fact that you choose to demo Whisper with this clunky example is strange. I’m sure it’s much more capable than this. I speak to my Echo dot and Google Home mini fluently. Whisper is supposed to be equally capable.