Tone Recognition Database of Electronic Pipe Organ Based on Artificial Intelligence

In the past few decades, artificial intelligence technology has experienced rapid development, and its application in modern industrial systems grown rapidly. This research mainly discusses construction of a database electronic pipe organ tone recognition based on intelligence. The timbre synthesis module realizes according to current parameters. audio time domain information (that is, data obtained by file analysis) is framed windowed, fast Fourier transform (FFT) performed each frame obtain frequency frame. harmonic peak method improved confidence used identify pitch, fundamental tone, calculate multiplier. Based parameters parameter editing interface, synthesized frame, then perform inverse waveform frame; connect waveforms different frames cross-average time-domain tone). After collecting sound organ, needs be denoised, imported parsed information. Then, are frequency-converted characteristic analyzed; through human-computer interaction interface intelligence, generated. If effect not satisfactory, you can re-edit generate timbre. During experiment, overall rate 3762 notes 286 beats was 88.6%. model designed this study flexibly libraries qualities meet requirements authenticity.