|
|
--- |
|
|
title: TextCut |
|
|
emoji: ✂️ |
|
|
colorFrom: blue |
|
|
colorTo: purple |
|
|
sdk: gradio |
|
|
sdk_version: 6.3.0 |
|
|
app_file: app.py |
|
|
pinned: false |
|
|
--- |
|
|
|
|
|
# TextCut |
|
|
|
|
|
Edit videos by simply editing their transcript. Upload a video, get an automatic transcription with timestamps using VibeVoice-ASR, then delete lines from the transcript to cut those parts from your video. |
|
|
|
|
|
## Features |
|
|
|
|
|
- **Automatic Transcription**: Uses Microsoft's VibeVoice-ASR model for accurate speech-to-text with timestamps |
|
|
- **Real-time Highlighting**: Current sentence is highlighted (uppercased) as the video plays |
|
|
- **Simple Editing**: Delete lines from the transcript to mark segments for removal |
|
|
- **Video Cutting**: Automatically cuts the video based on deleted transcript segments using FFmpeg |
|
|
|
|
|
## Usage |
|
|
|
|
|
1. **Upload**: Upload a video file (mp4, mov, etc.) |
|
|
2. **Transcribe**: Click "Transcribe" to generate the transcript with timestamps |
|
|
3. **Edit**: Delete lines from the transcript that you want to cut from the video |
|
|
4. **Apply Cuts**: Click "Apply Cuts" to generate the edited video |
|
|
|
|
|
## Requirements |
|
|
|
|
|
- Python 3.10+ |
|
|
- FFmpeg installed on the system |
|
|
- CUDA-capable GPU (for transcription) |
|
|
|
|
|
## Local Development |
|
|
|
|
|
```bash |
|
|
pip install -r requirements.txt |
|
|
python app.py |
|
|
``` |
|
|
|
|
|
## Hugging Face Spaces |
|
|
|
|
|
This app is designed to run on Hugging Face Spaces with ZeroGPU support for the transcription model. |
|
|
|
