Add new feature for accent detection.

[Appy-arsnel]

Hi Guys,
I’d like to propose adding a new feature for accent detection to OpenAI Whisper. While Whisper is already excellent at transcription and language detection, integrating accent detection would significantly enhance its capabilities for applications in language learning, accessibility, sociolinguistic research, and personalized AI experiences.

As a reference, I came across this model on Hugging Face that performs accent identification on English audio using a fine-tuned XLS-R model:
🔗 https://huggingface.co/Jzuluaga/accent-id-commonaccent_xlsr-en-english