Music Meets Artificial Intelligence

Speaker:
Andreas Arzt, Matthias Dorfer
Abstract:
Music is a fascinating domain and offers many opportunities for developing, testing, and applying signal processing and machine learning techniques. In our talk we will focus on selected tasks involving sheet music and music performances, such as automatic sheet music following for musicians and retrieving sheet music for recordings. We will cover ‘classic’ algorithmic approaches, robust enough to follow orchestral music at the Concertgebouw, as well as cutting-edge deep (reinforcement) learning methods that can be used to design intelligent agents that learn to read notes from sheet music images, listen to music, and follow music performances directly in the score image. The rich application potential, which we will demonstrate in concrete application scenarios, makes intelligent music processing a very active research field. We hope that the talk will serve as an inspiration for the audience to regard to music as a promising (and beautiful) object of study.
Length:
01:22:19
Date:
26/04/2018
views: 867

Images:
Attachments: (video, slides, etc.)
114 MB
491 downloads
892 MB
868 downloads
174 MB
446 downloads
255 MB
485 downloads
512 MB
481 downloads