https://github.com/labrijisaad/youtube-video-transcriptor

In this notebook, I implemented a script to transcribe YouTube videos (and audio files in general) using Google's speech-to-text API.
https://github.com/labrijisaad/youtube-video-transcriptor

audio-split general-idea-summarization googletranslateapi puthon speech-recognition speech-to-text text-translation transcript youtube youtube-transcripts youtube-video

Last synced: 3 months ago
JSON representation

In this notebook, I implemented a script to transcribe YouTube videos (and audio files in general) using Google's speech-to-text API.

• Host: GitHub
• URL: https://github.com/labrijisaad/youtube-video-transcriptor
• Owner: labrijisaad
• Created: 2022-07-26T15:48:11.000Z (over 2 years ago)
• Default Branch: main
• Last Pushed: 2022-12-19T09:15:12.000Z (about 2 years ago)
• Last Synced: 2024-10-10T00:43:43.055Z (3 months ago)
• Topics: audio-split, general-idea-summarization, googletranslateapi, puthon, speech-recognition, speech-to-text, text-translation, transcript, youtube, youtube-transcripts, youtube-video
• Language: Jupyter Notebook
• Homepage:
• Size: 58.6 KB
• Stars: 10
• Watchers: 1
• Forks: 2
• Open Issues: 2
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
## 🎥 `Youtube-video-transcriptor in Python` 🐍



  



In this project, I developed a script in Python that uses Google's speech-to-text technology to transcribe audio from YouTube videos.

- ⚠️ Please note the following before using the script:

>   - 1️⃣ **`The script is intended to be run on Google Colaboratory!`**  

>   - 2️⃣ The script may not always accurately transcribe text due to noise or the way the speaker talks in the video (e.g. speaking too fast or too slow).

>   - 3️⃣ **`The summary model`** used in the script is a community model available on [Huggingface](https://huggingface.co/) that **`only supports English text`**. It may not always accurately capture the general idea of the transcription, especially if there is a lack of data.

 

❓❓❓ HOW TO USE ❓❓❓


```

>>> 1️⃣ Run the notebook in Colab (make sure you are logged into Colab with your Google account).

>>> 2️⃣ Paste the URL of the youtube video you want to transcribe into the `url` variable.

>>> 3️⃣ Replace the `lang` variable with the language spoken in the video (all instructions are provided in the notebook).

>>> 4️⃣ Run all cells (shortcut: `CTRL + F9`)

>>> 5️⃣ Download the generated TXT files (there will be two in total: one for the transcription and one for the translated transcription).

```

⚠️⚠️⚠️ UPDATE ⚠️⚠️⚠️


```

>>> To optimize transcription time, I have updated the script to use `Python threads`, which helps to fully utilize the CPU resources provided by Colab. 

>>> As a result, the performance has significantly improved - a 30-minute video can now be transcribed in approximately 35 seconds, compared to the previous time of 2 minutes and 30 seconds. 

>>> You can find the updated script with threads in the accompanying notebook. 😁

```

> - 🙌 Notebook made by [@labriji_saad](https://github.com/labrijisaad)

> - 🔗 Linledin [@labriji_saad](https://www.linkedin.com/in/labrijisaad/)

- 📫 Feel free to contact me if anything is wrong or if anything needs to be changed 😎!  **[email protected]**