TuneFinder: Search for Music with Your Humming

Object Type: Asset
In Folder: TuneFinder

Expand to show object metadata

TuneFinder: Search for Music with Your Humming

Senior Design Showcase project, Wentworth Institute of Technology, College of Engineering and Computer Science, Department of Computer Science and Networking, 2020

Dublin Core Metadata

title: TuneFinder: Search for Music with Your Humming

creator: Eric Scherfling, Nick Thai

subject: Machine learning, Music, Web application

description: Songs are made up of notes and often lyrics to accompany the notes. Often time people can remember the lyrics and notes, but not the name of the song. Although Shazam-like applications are extremely powerful, they normally match clips of songs to the songs. The problem comes when people attempt to sing the songs; they may not end up singing in the same key, or at the same tempo as the recorded song. Solving this problem would require some computations to be done on a song to reduce it down to notes that could be compared to someone else’s sings. In this project, we attempt to address identified problem by using machine learning at the back-end infrastructure and web programming skills for the front-end user interface. There are many research topics occurring in this project, including music encoding and model training, input data and formatting, model prediction and result presentation. In terms of implementation, there are also many technical issues. For example, .wav format deals with what frequencies are present at a certain point in time, but simply by looking at these frequencies, notes can be estimated. Although programs like WAoN help turn Wav files into Midi format (a format which directly deals with notes and times), they are not perfect and are vulnerable to background noise in Wav files. The result of our project is a web-based service with a customized algorithm that can allow the users to search for a song with the user’s humming as an .wav input.

publisher: The Douglas D. Schumann Library and Learning Commons, Wentworth Institute of Technology

contributor:

date: 2020

type: Poster

format: .pdf

identifier:

source:

language: eng

relation:

coverage:

rights: https://rightsstatements.org/vocab/InC/1.0/

Show more ▼