Assemble AI Challenge

Sophisticated Speech-to-Text

What I Built

I created a sophisticated Speech-to-Text application powered by AssemblyAI’s Universal-2 model. My project not only transcribes audio but also incorporates advanced features to enhance user experience.

Key Features:

• Accurate Transcription: Leveraging AssemblyAI’s Universal-2 model for precise transcription of audio files.
• Speaker Statistics: Detailed insights into speaker activity, including the number of speakers, speaking time, and word counts.
• Synchronous Audio Playback: Audio playback synchronized with the transcribed text, providing a seamless experience for reviewing transcripts.
• Export Options: Ability to export transcriptions in multiple formats, including .txt files.
• User-Friendly Interface: Intuitive design to interact with transcriptions and features efficiently.

Demo

• GitHub Repository: https://github.com/Gopinathv19/AssembleAI-Challenge2024

• Demo Video: https://www.youtube.com/watch?v=kY4BvFr-Log

Screenshots and visuals of the app can be found in the GitHub repository.

Journey

This project integrates AssemblyAI’s Universal-2 Speech-to-Text Model to deliver accurate and reliable transcription. I built upon this core functionality to add value through unique features like speaker statistics and synchronized audio-text playback, which are not commonly found in basic transcription tools. These features aim to provide a better user experience and ensure the application goes beyond simple transcription.

I also focused on export functionality, enabling users to download transcriptions in .txt and potentially other formats, making it versatile for various use cases.

As a solo participant, I handled all aspects of the project, from design and implementation to testing and optimization. It was a rewarding experience to push the boundaries of what a transcription tool can do while ensuring the application remains user-friendly and efficient.

Thank you for considering my submission!