The Extended Kalman Filter is used to track fundamental frequency, amplitude and instantaneous phase of audio signals. This method is an addition to the extensive pre-existing literature available on pitch detection. It has certain advantages including the fact that it gives a unique pitch value for each sample of data, unlike most block-based methods like cepstrum or YIN estimator, and is robust to the presence of a large amount of observation noise. However, it has certain drawbacks such as poor transient performance and slow detection of rapid pitch changes.
This paper has received the 4th best paper award at DAFx 2017 and an extended version has been invited for publication in the Journal of the Audio Engineering Society (JAES).
Artist Identification by analyzing musical score is a unique problem in MIR. In this project, we tackle the novel problem of automatically classifying guitarists by analyzing guitar solo tabs. We have created a hand-curated dataset of 80 guitar solos downloaded in MusicXML format played by Eric Clapton, David Gilmour, Mark Knopfler and Jimi Hendrix. The solos have been cleaned and converted to tuples containing note duration, transposed fret and string. We have tried to visualize the stylistic difference between artists by computing pitch class and beat histograms and self-similarity matrices. As an initial classification paradigm, we have used zero and first-order Markov chains, the latter giving significantly above chance classification accuracy. Work is still in progress and we wish to employ some modern Machine Learning techniques for better classification result.
The Ranchlands' Hum is a low frequency noise around 40Hz that has been plaguing the residents of Calgary, Canada for years. As an intern in the department of Electrical and Computer Engineering at University of Calgary, I was assisting Dr. Mike Smith in developing an Android application that could capture, store and analyze low frequency noise. I added features that integrated the existing application with an SQLite database, calculated and plotted signal metrics. The project received some media attention.
The Kalman Filter is an MMSE estimator that can be used to remove background noise from speech. The filter equations are formulated based on the linear Autoregressive model of speech production. I implemented a novel algorithm that tuned the Kalman Filter by accurately determining its parameters - measurement and process noise covariance. I also studied the effect of changing AR model order on speech corrupted with various types of noise of various SNRs and concluded the results in my undergraduate thesis.
Tãbla is is a membranophone percussion instrument (similar to bongos) which is often used in Hindustani classical music. The instrument consists of a pair of hand drums of contrasting sizes and timbres. The rhythmic pattern of any composition in Indian music is described by the term tãla, which is composed of cycles of mãtra-s. Tãla roughly correlates with the metres in Western music. Our aim was to determine the number of beats that constituted tãla-s in different tãbla solos. We developed a heuristic algorithm that extracted peaks from the tãbla signal, corresponding to single or composite strokes and devised statistical methods to ensure that spurious noisy peaks were removed,and missed peaks were accounted for. We obtained excellent results for solo tãbla recordings played by human artist.