https://github.com/locaal-ai/video-transcript-helper
Use transcription to Zap filler words (uh's & um's), generate content e.g. blog post, YouTube description
https://github.com/locaal-ai/video-transcript-helper
ai llm speech-to-text
Last synced: 7 days ago
JSON representation
Use transcription to Zap filler words (uh's & um's), generate content e.g. blog post, YouTube description
- Host: GitHub
- URL: https://github.com/locaal-ai/video-transcript-helper
- Owner: locaal-ai
- License: mit
- Archived: true
- Created: 2023-05-17T15:48:10.000Z (almost 2 years ago)
- Default Branch: main
- Last Pushed: 2024-01-26T02:50:12.000Z (over 1 year ago)
- Last Synced: 2025-03-29T05:23:01.935Z (about 1 month ago)
- Topics: ai, llm, speech-to-text
- Language: Python
- Homepage:
- Size: 39.1 KB
- Stars: 10
- Watchers: 2
- Forks: 4
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Video Transcript Helper
[](https://discord.gg/KbjGU2vvUz)
A comprehensive toolkit designed for content creators, educators, digital marketers, and video editing enthusiasts.
It harnesses the power of AI and video processing through a suite of Python scripts that simplify the post-production process.
This free open-source project aims to transform the way users handle video content, turning hours of editing into a task of a few command lines.
This project contains three scripts:
- `transcribe_from_video_XXX.py`: Transcribe a video
- `clean_video_from_transcription.py`: Zap filler words ('uh', 'um') in videos using FFMPEG
- `summary_chapters_blog.py`: Generate a summary, video chapters and a blog post
Roadmap of future features:
- Remove or speedup (shorten) periods of "silence"
- Enhance speech by voice separation models
- Generate a supercut for a quick video snippet
- Add Audiogram / Kareoke kind of subtitles on the video
- Translate the subtitles to any language
## Usage
Transcribe the video: (either AWS Transcribe API or [Faster-Whisper](https://github.com/guillaumekln/faster-whisper))
```sh
$ python transcribe_from_video_XXX.py
```
The output will be a file called `.json` in the same directory as the video.
Zap the filler words:
```sh
$ python clean_video_from_transcription.py
```
The output will be a file called `-clean.mp4` in the same directory as the video.
Generate the summary, chapters and blog post:
```sh
$ python summary_chapters_blog.py --generate_summary --generate_chapters --generate_blog
```
## Dependencies
- Python 3.6+
- [FFMPEG](https://ffmpeg.org/)
- [AWS CLI](https://aws.amazon.com/cli/)
Make sure to configure your AWS CLI with your credentials and region.