readme.md

metadata

title: Accent Classifier
emoji: 🎙️
colorFrom: teal
colorTo: cyan
sdk: gradio
sdk_version: 3.38.1
app_file: app.py
pinned: false

Accent Classifier 🎙️

This app downloads a public YouTube or Vimeo video, extracts its audio, and classifies the speaker's accent (or rather, speaker ID as a proxy) using a Hugging Face model.

How it works

1. You provide a video URL.
2. The app downloads the audio using yt-dlp.
3. It extracts the audio in a format suitable for the model (wav, 16kHz, mono).
4. It runs the superb/wav2vec2-base-superb-sid model to classify the speaker.
5. Displays the predicted speaker ID and confidence.

Requirements

• Python 3.8+
• yt-dlp
• ffmpeg installed on your system and accessible from the command line.
• gradio for the UI.
• transformers from Hugging Face.

Usage

Run the app:

python app.py