(Won first place in our track in the hackathon)
Deployed Link: Speech-Diarization

Problem Statement:

1. Multi-speaker audio recordings are difficult to analyze.
2. Manual transcription is time-consuming and error-prone.
3. Need for automated speaker identification and transcription.

Proposed Solution

1. A Streamlit-based web app for speaker separation and transcription.
2. Uses AssemblyAI for accurate transcription with speaker labels.
3. Provides speaker statistics for better analysis.

Future Improvement

1. Support for more languages.
2. Noise reduction for noisy environments.
3. Real-time transcription during live calls.