Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/sofianedjerbi/brainfryer
Generate attention-grabbing videos!
https://github.com/sofianedjerbi/brainfryer
Last synced: about 2 months ago
JSON representation
Generate attention-grabbing videos!
- Host: GitHub
- URL: https://github.com/sofianedjerbi/brainfryer
- Owner: sofianedjerbi
- License: gpl-3.0
- Created: 2024-03-14T12:50:59.000Z (9 months ago)
- Default Branch: main
- Last Pushed: 2024-03-20T13:56:50.000Z (9 months ago)
- Last Synced: 2024-10-12T03:44:21.314Z (2 months ago)
- Language: Python
- Size: 103 MB
- Stars: 1
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
BrainFryer
Turn online discussions into captivating videos!
Join our creative community!
[](LICENSE)
[](https://discord.gg/jwb26Xy5M7)
[](https://www.tiktok.com/@redditeel)
## How to use
### Basic Usage
You can use the wizard by typing `python -m brainfryer`.
A CLI is available: `python -m brainfryer `
### CLI Options
Brainfryer supports several options for customization:
- `-h`, `--help`: Display the help message and exit.
- `-c COMMENTS`, `--comments COMMENTS`: Specify the maximum number of comments to retrieve (default is 10).
- `-b BACKGROUND`, `--background BACKGROUND`: Provide a YouTube URL for the video background (optional).
- `-s SONG`, `--song SONG`: Provide a YouTube URL for the background song (optional).
- `-i`, `--gen-images`: Enable image generation for your video (disabled by default).
- `-t`, `--gen-subtitles`: Enable subtitle generation for your video (disabled by default).
- `-d`, `--dall-e`: Use Dall-E for image generation instead of Stable Diffusion (disabled by default).
- `-u`, `--upload`: Upload the generated video on TikTok using `cookies.txt` (disabled by default).
### Example
To generate a video with images, subtitles, and upload it to TikTok, using a specific Reddit URL, background, and song, your command might look like this:
```bash
python -m brainfryer https://www.reddit.com/r/exampleSubreddit/examplePost -c 20 -b https://youtu.be/exampleBackground -s https://youtu.be/exampleSong -i -t -u
```
This command uses a Reddit URL, sets the maximum comments to 20, specifies background and song YouTube URLs, enables image and subtitle generation, and uploads the final product to TikTok.
### Demo
A live demo is available [here](https://www.tiktok.com/@redditeel).
## Setup
BrainFryer uses OpenAI's **TTS API** and **ChatGPT** for illustrations.
- Clone the repository: `https://github.com/chaoxel/BrainFryer.git`
- Enter the directory: `cd BrainFryer`
- Install dependencies: `pip install -r requirements.txt`
- Install [playwright](https://playwright.dev/python/docs/intro): `playwright install`
- Copy `.env.example`: `cp .env.example .env`
- Configure `.env`: `nano .env`
If you need subtitles:
- Install [CUDA Toolkit](https://developer.nvidia.com/cuda-downloads)
- Install [cuDNN](https://developer.nvidia.com/cudnn)
If you need illustrations:
- With [Dall-E](https://openai.com/dall-e-3): No configuration needed.
- With [Stable Diffusion](https://github.com/AUTOMATIC1111/stable-diffusion-webui): Install the [webui](https://github.com/AUTOMATIC1111/stable-diffusion-webui) and run it with `--api`. Configure *host* and *port* in `.env`.
## Behind the Scenes
Brainfryer turns Reddit threads into videos through a multi-step process:
1. **Webpage Parsing**: Utilizes Playwright to navigate and parse the target webpage for content extraction.
2. **Screenshot Capture**: Takes screenshots of the webpage, capturing key visual elements.
3. **Content Analysis**: Uses ChatGPT to analyze the extracted content and identify themes or key points.
4. **Illustration Prompt Generation**: Based on the analysis, generates creative prompts for illustrations.
5. **Image Generation**: Sends these prompts to Stable Diffusion or Dall-E, which then generate corresponding images.
6. **Text-to-Speech Synthesis**: Converts the extracted textual content into spoken words using OpenAI's TTS API.
7. **Background Video Processing**:
- Downloads the specified YouTube background video using Pytube.
- Edits and crops the background video to fit the final video's format using Moviepy.
8. **Video Assembly**: Combines the generated images, edited background video, and synthesized speech into a cohesive video using Moviepy.
9. **Audio Transcription**: Converts the final video's audio into text using Whisper for subtitle generation.
10. **Subtitle Timing**: Analyzes the transcription to add accurate timestamps for each word, creating synchronized subtitles.
11. **Final Editing**: Applies the subtitles to the video, adjusting for readability and style.