Create README.md

# Accent Classification from English Speech in Video for PRE Interview

This project uses a pretrained ECAPA-TDNN model from [Hugging Face](https://huggingface.co/Jzuluaga/accent-id-commonaccent_ecapa) to classify **English accents** (e.g., British, Indian, American) from **spoken audio in `.mp4` video files**. It extracts audio, resamples it, and runs classification using the `speechbrain` library.

---

## 📦Dependencies

Install required Python libraries:

# macOS
brew install ffmpeg

# Ubuntu
sudo apt install ffmpeg

```bash
pip install torch torchaudio librosa pydub soundfile speechbrain ffmpeg-python

## Implementation Note

I originally experimented with downloading videos directly from **YouTube** using automated tools such as [`yt-dlp`](https://github.com/yt-dlp/yt-dlp). However, this approach required users to:

- Manually authenticate with YouTube
- Download session cookies frequently

This broke the automation goal and added friction for non-technical users.

**So I decided to simplify the experience by requiring users to manually provide the `.mp4` file.**

README.md

@@ -0,0 +1,9 @@
+---
+license: mit
+title: English Accent Detector Parser
+sdk: gradio
+emoji: 🏢
+colorFrom: indigo
+colorTo: green
+short_description: PreInterview project
+---