S.A. Abdallah and M.D. Plumbley. Sparse coding of music signals. available on-line at http://www.kcl.ac.uk/depsta/pse/diveng/eleceng/audio/index.html, 2001.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....noisy sparseness. 1. INTRODUCTION In image analysis and vision research, sparseness has been demonstrated to be a powerful concept in finding meaningful representations of data [4, 11, 12, 3, 6, 7, 17, 15, 10] The concept of sparseness or sparsity is also used in speech and music analysis [9, 2], in the statistical modeling of natural languages [16] and in various other applications. Despite the popularity of the sparse ideology , sparseness is not unambiguously defined. The simplest definition of sparseness states that in a sparse matrix or vector most of elements are zero. This ....

S. Abdallah and M. Plumbley. Sparse coding of music signals, submitted for publication, 2001.

....and or instruments being played at the same time, and we would like to decompose this into the note characteristics, and when each note is being played. In previous work we have approached this problem in the frequency spectrum domain, using Saund s [1] Multiple Cause Model [2] and sparse coding [3]. In its simplest form, this type of problem, in common with many others, can be considered to be a non negative factor analysis problem. Each note has a positive (or zero) volume, and the playing of a given note contributes (approximately) a positive amount of power to a given frequency band. ....

S. A. Abdallah and M. D. Plumbley, "Sparse coding of music signals," Submitted to Neural Computation, 2001.

....and or instruments being played at the same time, and we would like to decompose this into the note characteristics, and when each note is being played. In previous work we have approached this problem in the frequency spectrum domain, using Saund s [1] Multiple Cause Model [2] and sparse coding [3]. In its simplest form, this type of problem, in common with many others, can be considered to be a non negative factor analysis problem. Each note has a positive (or zero) volume, and the playing of a given note contributes (approximately) a positive amount of power to a given frequency band. ....

S. A. Abdallah and M. D. Plumbley, "Sparse coding of music signals," Submitted to Neural Computation, 2001.

....The observation in the previous section that a transcription of a musical sound has a sparse represenation suggests that sparse coding could be used directly in the search for a possible transcription. We have applied this sparse coding approach to the analysis of polyphonic harpsichord music [1]. We use a prior for note probability that contains a sharp peak at zero. This leads to a coding of the input spectra where most notes are o , and only a few notes represent the input. We tested our algorithm on real sounds generated by a MIDI synthesizer and then resampled. We used Bach s ....

....resampled. We used Bach s Partita in A minor for keyboard, BWV827, which was chosen since it mainly consists of two or three independent lines, with a few block chords. We used magnitude spectra rather than power spectra so that the noise be more approximately Gaussian, as assumed in our model [1]. The model learned 55 basic spectra, 49 of which were note spectra, and others appeared to represent transistions. When the output activations of these basis vectors are plotted, they show a good pianoroll representation for the music contained in the input spectrogram (Fig. 5) To check, a ....

S. A. Abdallah and M. D. Plumbley. Sparse coding of music signals. Submitted for publication, 2001.

No context found.

S.A. Abdallah and M.D. Plumbley. Sparse coding of music signals. available on-line at http://www.kcl.ac.uk/depsta/pse/diveng/eleceng/audio/index.html, 2001.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC