Live Speech to Text with Watson Speech to Text and Python | FREE Speech to Text API

Want to skip out on copying down lecture notes?

Maybe you want a live transcript from a meeting?

To do that, you can use live speech to text transcription. In this video you'll learn how to do exactly that.

We'll use the FREE tier of the Watson Speech to Text service to transcribe audio from your microphone live. You'll be able to see the transcription as it's being spoke. And better yet, you can leverage the pre-written code to do it!

In this video, you'll learn how to:
1. Setup Watson Speech to Text for FREE
2. Access the Live Speech to Text Code
3. Perform live Speech to Text Transcription using your Microphone

Chapters:
0:00​ - Start
2:38 - Setting up Watson Speech to Text
6:22 - Accessing the Streaming Speech to Text code from GitHub
8:34 - Installing dependencies
12:22 - Setting up Configuration
16:34 - Running Live Speech to Text

Oh, and don't forget to connect with me!

Happy coding!
Nick

P.s. Let me know how you go and drop a comment if you need a hand!

[StevePGLy]

Hey Nicholas. I am Steve who is on a learning journey to become data scientist. I would like to say that I really appreciate what you are sharing! I have been studying pretty much by watching your great videos :D

[thinamG]

This looks awesome, Nicholas. Thanks for sharing!

[ravindunawanjana]

I already started to followed your videos and this is great video nick cheers

[guyincognito]

Thanks for zooming in to make the text larger, and for using a larger font in VS Code. Also thanks for alerting me to pipwin!

[TimDownsAnimation]

I'm an amateur at this stuff, but I'm trying to design a proof-of-concept that proposes a hybrid of something like this with lip-reading AI to produce real-time subtitles for the deaf and hard-of-hearing. In my research, I came across your channel and I love it! You're great at explaining things and it's easy to follow along. Instant sub!

[ComputerScienceSimplified]

Amazing video, keep up the incredible work! :)

[haisomeone]

You don't know or somone didn't tell you that you are the best.

[valeriofaraone]

hey nicholas thank you very much, you are great.

[avnishat]

@Nicholas You specified this can be used for live transcript from a meeting but when I tried this transcribes only the voice from my microphone and does not transcribe others voice from the meeting so how it can be used for transcribing a meeting. I would love to use this code to transcribe the entire meeting session

[pauljones]

Hi Nicholas,

I've watched a few videos on how to use speech to text, with the majority being with google. I must say your explantion is more detailed, and the audio quality is very good (crisp and clear). It seems like watson is easier to install than googles version. I'd like to install watson on my macbook. Is there anything that I need to do differently, or can I follow the exact steps you used to get it installed on my mac?

Will watson detect speech if I have a video playing on my mac and then transcribe that audio to text?

Kind Regards
Paul

[WisKyVT]

Nice! what if you wanted to have it automatically trigger when someone starts talking, and stops after like 2 seconds of silence?

[mailtoraj]

Great work!!, but can I get the final text somewhere into the file preferably in JSON? I need to pass this to my app. Advise pls.

[viniciuslongo]

Hands down the best ML youtube channel

[adibokay]

This is so helpful, thank you Nick! I am trying to build a software that can convert ASL to text on the screen that can help ease communication on platforms such as Zoom and Microsoft Teams. Would be great if you could show us how to build it ^^

[oluwatimilehinfolarin]

Thank you Nicholas for the wonderful video.

But I got an error - [Errno 11001] getaddrinfo failed.

How can I solve this problem? It's urgent. Thank you.

[OliverHiggins]

Great video. I got the 403 issues to (Based in Sydney) the only location that would work was us-south

[shoaqa]

amazing video, thank you
i'm having a problem with the last text is not being printed and idk why
i added a print statement to on_close to see if it reaches it but it didn't. even though it was working fine before then it stopped so idk what to do :(

[lenover]

this looks really awesome, I was wondering if it was possible to output the phonemes instead of text. that would be something I could really use in a project!

[razmandhamarasheed]

Hey man, great work! I love your explanation. Want to get into data science and learning it from you makes it easy. I have a request though, are you able/willing to look more into process time series data where you can make continuous predictions based on historical data?

Anyways, keep up the great work!

[ajkadhim]

thank you so much for an excellent video. When it is transcribing live, it outputs several lines as you are speaking. However, after "done recording", you see the complete output of the text. When I am running the command, I do not have that finalized output at the end. Is there a way to get that? thank you again in advance.