Pitch Detection in Singing Evaluation (Scoring) in Karaoke using Pyaudio, Pygame and VPython

Ren-yuan Lyu

Audience level:
Intermediate
Category:
Science

Description

Based on digital signal processing (DSP) theory, using the Pyaudio, Pygame and VPython modules, a real time pitch detection module was applied to a 3D scenary to control objects in the scene. It can be used as a kernel module in a singing evaluation system similar to the Karaoke machine used in the Karaoke Battle (カラオケバトル).

Abstract

Introduction ==== This project was implemented in the Python 3.4 platform, using multi-threading programming technique and scientific computation modules in a simple manner. It was based on a 2014 project, which dealt with the FFT spectrogram of an audio signal. This year, the author extended it into audio feature extraction, especially for the pitch detection in human singing signal. A popular application of such a system is in the scoring of a human singing, e.g., the machine used in the KaraOke Battle (カラオケバトル)TV program. The Task ==== ![](https://dl.dropboxusercontent.com/u/33089565/_pyconjp2016/1001_theTask.png) The Prototype ==== ![](https://dl.dropboxusercontent.com/u/33089565/_pyconjp2016/1002_thePrototype.png) The Modules ==== ![](https://dl.dropboxusercontent.com/u/33089565/_pyconjp2016/1003_theModules.png)
  • このエントリーをはてなブックマークに追加