Skip to content

Latest commit

 

History

History
43 lines (30 loc) · 1.73 KB

README.md

File metadata and controls

43 lines (30 loc) · 1.73 KB
Raw

Hackday Project

Idea

TrackMeaning is an application that takes as an input a track (WAV file) and a prefferred language (english, french, german, etc) and it gives you back the message of the song (what is the artist singing about) in the language you specified.

Steps

  1. Get the id of the track of the specified track using Recognise Audio API
  2. Get the lyrics of the song using Lookup Track API
  3. Translate the lyrics in the preffered language using translation API
  4. Determine the topic of the text using NLP techniques
  5. Present the topic to the user

Technologies

Backend

  • Programming language - Go
  • Framework for creating the REST API - Gin
  • External APIs used - Recognition Audio API, Lookup Track API, Google NLP API, Google Translate API

Frontend

  • Programming language - JavaScript
  • Library - React, Redux
  • Other technologies for building and compiling JS - Webpack, Babel

Improvements

  • Most of the http requests can be turned into Goroutines so that they run concurrently
  • Improve the NLP techniques used so the result is better
  • Add more languages (currently only four are available - English, French, Italian, Bulgarian)

Build

To run the application:

  1. Get SHAZAM API (.txt) and GOOGLE API credentials (.json) that you need to place inside /api/go as wel
  2. Run from the main directory docker build --rm -f api/react/Dockerfile -t boyan.bonev:intern-hackday-react api/react
  3. Run from the main directory docker build --rm -f api/go/Dockerfile -t boyan.bonev:intern-hackday-go api/go
  4. Run docker run -d -p 5000:5000 boyan.bonev:intern-hackday-react
  5. Run docker run -d -p 8080:8080 boyan.bonev:intern-hackday-go

Go to localhost:5000 and you will see the app running.