Applications of Stochastic Machine Learning and Statistical Signal Processing Approaches to Automatic Music Transcription and Visualisation

点击翻译按钮获取中文摘要

The process of transcribing complex polyphonic and polytimbral performances to musical notation is a notoriously arduous task, requiring the ability to distinguish numerous instrumental lines, separable often only via remarkably subtle variances in timbre, frequency, and waveform characteristics. The proposed research involves investigating several novel techniques for the automation of the transcription process, combined with the exploration of automated visualisation utilities derived through deep learning techniques, and analysis of the limitations of previously derived methods.