Supervised feature learning via sparse coding for music information rerieval
O'Brien, Cian John
MetadataShow full item record
This thesis explores the ideas of feature learning and sparse coding for Music Information Retrieval (MIR). Sparse coding is an algorithm which aims to learn new feature representations from data automatically. In contrast to previous work which uses sparse coding in an MIR context the concept of supervised sparse coding is also investigated, which makes use of the ground-truth labels explicitly during the learning process. Here sparse coding and supervised coding are applied to two MIR problems: classification of musical genre and recognition of the emotional content of music. A variation of Label Consistent K-SVD is used to add supervision during the dictionary learning process. In the case of Music Genre Recognition (MGR) an additional discriminative term is added to encourage tracks from the same genre to have similar sparse codes. For Music Emotion Recognition (MER) a linear regression term is added to learn an optimal classifier and dictionary pair. These results indicate that while sparse coding performs well for MGR, the additional supervision fails to improve the performance. In the case of MER, supervised coding significantly outperforms both standard sparse coding and commonly used designed features, namely MFCC and pitch chroma.
Showing items related by title, author, creator and subject.
Meinz, Elizabeth J. (Georgia Institute of Technology, 1996-05)
Tsuchiya, Takahiko; Freeman, Jason; Lerner, Lee W. (Georgia Institute of TechnologyUniversity of Music and Performing Arts Graz. Institute of Electronic Music and Acoustics, 2015-07)In sonification methodologies that aim to represent the underlying data accurately, musical or artistic approaches are often dismissed as being not transparent, likely to distort the data, not generalizable, or not ...
Gandy, Maribeth; Quay, Andrew (Georgia Institute of TechnologyInternational Community for Auditory Display, 2000-04)The Haptic Music system allows up to three people to collaboratively create music in a jazz style. The users can have varying musical backgrounds, and even a novice can use his/her creativity to create a novel composition. ...