Vanhack English Pre-Test - Backend

This is was a submission to Vanhack for vanhackathon 2019

Overview

Install and run

For instructions for the Frontend, please click here.

For the Backend, follow these steps: You need to have pip3/python 3 installed

git clone https://github.com/willianbs/vhk-backend-speech.git

Go to the project directory and run:

python3 main.py

The API endpoint is [POST] http://localhost:5000/file-upload to upload a MP4 file.

Design and Architecture

Frontend App

A lightweight React.js app

Backend API

The backend api was implemented as a Python simple server with Flask to deal with the upload request and the speech analysis. No login required at this time.

About the project

Problem

One important part of VanHack process is the English verification process. It needs a real person to evaluate all entries and it can take a lot of time of a small team.

Solution

My idea can go 2 ways (or both):

Using as a tool for the users (Premium only?) to test their proficiency before submitting each video. So they'll know if they've performed well enough to get a good grade.
As an auto evaluation system for the Staff to reject/approve based on AI and if the machine isn't able to get it right or there's a claim, the Staff can optimize the work.

And that's about it :)

Analysis based on: https://github.com/Shahabks/my-voice-analysis