Browsing School of Electronic Engineering and Computer Science by Author "STURM, BLT"
Now showing items 1-8 of 8
-
Deep Learning and Music Adversaries
STURM, BLT; Kereliuk, C; Larsen, J (IEEE Explore, 2015-09-10)An {\em adversary} is essentially an algorithm intent on making a classification system perform in some particular way given an input, e.g., increase the probability of a false negative. Recent work builds adversaries for ... -
?`El Caballo Viejo? Latin Genre Recognition with Deep Learning and Spectral Periodicity
STURM, BLT; Kereliuk, C; Larsen, J; Mathematics and Computation in MusicThe ``winning'' system in the 2013 MIREX Latin Genre Classification Task was a deep neural network trained with simple features. An explanation for its winning performance has yet to be found. In previous work, we built ... -
``Horse'' Inside: Seeking Causes of the Behaviours of Music Content Analysis Systems
STURM, BLT (ACM, 2016-12-16)Building systems that possess the sensitivity and intelligence to identify and describe high-level attributes in music audio signals continues to be an elusive goal, but one that surely has broad and deep implications for ... -
Music transcription modelling and composition using deep learning
STURM, BLT; Santos, J; Ben-Tal, O; Korshunova, I; 1st Conference on Computer Simulation of Musical Creativity -
Revisiting Priorities: Improving MIR Evaluation Practices
STURM, BLT; International Society for Music Information Retrieval -
The scientific evaluation of music content analysis systems: Valid empirical foundations for future real-world impact
STURM, BLT; Maruri-Aguilar, H; Parker, B; Grossmann, H; International Conference on Machine LearningWe discuss the problem of music content analysis within the formal framework of experimental design. -
Working Toward Computer-Augmented Music Traditions
STURM, BLT; Ben-Tal, O; DMRN+11: Digital Music Research Network One-day Workshop 2016 (2016-12-20)We discuss our work in modelling and generating music transcriptions using deep recurrent neural networks. In contrast to similar work, we focus on creating a rich evaluation methodology that seeks to address questions ...