Overview
In a world of diverse cultures, travelers face the challenge of cross-cultural language understanding. Language barriers can leave them feeling disconnected, relying on non-verbal cues to navigate conversations. Curiosity and the desire to bridge the gap intensify, yearning to unravel rich traditions and stories. This universal struggle reminds us that language holds the key to profound understanding, empathy, and meaningful connections. Embrace the beauty of cultural diversity as we break these barriers together with Transcribe Now Application.
Problem Statement
In today's globalized world, cross-cultural communication poses a significant challenge for travelers and individuals seeking to connect with people from different backgrounds. Language barriers hinder effective interaction, leading to missed opportunities for meaningful conversations, cultural understanding, and building connections.
Proposed Solution
Transcribe Now is a simple web application powered by advanced speech-to-text technology and machine learning algorithms,travelers can embark on their journeys with confidence, knowing they have a powerful tool at their fingertips.
They can easily communicate, ask questions, and engage in meaningful discussions, regardless of language differences. The app empowers users to unlock the full potential of cross-cultural experiences, fostering deeper connections, and facilitating genuine cultural exchange.
The highlight- Transcribe Now makes it easier than ever! Enjoy seamless communication without the need for authentication. Experience the freedom of instant cross-cultural understanding at your fingertips.
Tech Stack
NodeJs, React, Transformer tools
Our Approach
Transcribe Now has two approach
The Frontend
This is our user interface where the User get to see the homepage and on clicking "Get Started button" they are redirected to another page where they can record an audio in get the transcripted words displayed in real-time. It also has a start and stop button to control the recordings.
stage 1: collect recordings
The app uses browser MediaAPI to allow recordings while in the browser, By clicking the start button, the recordings start automatically and the user can have control of stopping it to.
stage 2: Generate chunks of audio file
Splitting it into chunks of audio file so it could be passed as a "blobUrl" to the backend for transcription.
stage 3: Pass to the Frontend
Created a function that accepts the generated chunks of audio file as an argument and passing it as a formdata, Then making a POST request to the specified url for transcription.
The Backend
stage 1: Endpoint creation
created an endpoint using Express.js server that handles a POST request and sends a request to an endpoint
stage 2 : Connect the Data from the frontend
Grab the url from the frontend "AudioBlob" through the API request body and converts to an acceptable format for transcription.
stage 3: Transcription
Transcribes the recordings using Whisper Transformer model and sends the transcripted words to the Frontend.
Outcomes and Learnings During the Process
It was not an easy Journey to be honest, Identifying the team members strengths and weakness and working on a project was one of our biggest challenge. We had to try out a couple of tech stack and AWS services to get the one to suit our interest.
Initially we tried using a Python environment for our backend by maximizing AWS lambda and API gateway and AWS Amplify to help us use the transformer tool but the file size was too large and then we tried Amazon Elastic File System(EFS) but we still could not get the app working.
Reading through the docs again after taking a lot of time trying to debug our code, we decided to go through with Nodejs because that was something we were more familiar with to build the Project.
Lessons
For my team, these are the couple of things we learnt while working on the project
Reading docs- we definitely read lots of them
Teamwork
showed up when we all tried to bring our strengths and skills to the projectBrainstorming
We did a lot of that when we tried getting a project idea, why we should with it and how well it suited the project Instructions
I am sure every member of the team must have learnt a couple of things from the hackathon experience
Way Forward
We hope to modify and refactor our codebase,To help people easily navigate through the codebase.
For the application, we have our users interest at heart and we definitely hope to integrate more language translation on it.
Top comments (1)
This appears to be a great addition. Welcome to the community.