Deep Learning for MIR II: State-of-the-art Algorithms

Tuition scholarships available for applicants commited to advancing diversity in STEM. Fill out this form by May 1st.

Instructors: Irán Román, Kitty Shi

A survey of cutting-edge recent research in MIR using Deep Learning. Students will build state-of-the art models using tensorflow* and GPU computing. Emphasis on generative models and reinforcement learning. Topics covered: music and speech synthesis, beat-tracking, music-recomendation, and semantic analysis. Students solve a real problem of their choice using state-of-the-art Deep Learning Models. This course is meant for individuals who want to gain experience applying Deep Learning to solve current problems in MIR. 

*Students may use other toolboxes if they are already familiar with them.

Prerequisites:
- Experience training Neural Networks or other machine learning algorithms
- Deep Learning for MIR I

About the instructors:
Irán Román studies Computer-based Music Theory at CCRMA and Neuroscience at the Stanford Neuroscience Institute. His dissertation models brain computations during sound perception using artificial neural networks. His industry experience includes the development of artificial neural networks for various applications.

Kitty Shi is a PhD Candidate at CCRMA, and a PhD minor in computer science. She is interested in modeling expressive musical performances. She's been working at Shazam, Adobe for various applications of MIR. 



IMPORTANT: Contact the instructor before registering to confirm your eligibility. Attach a copy of your registration or diploma for the CCRMA Deep Learning for MIR I workshop. Describe your experience with python programming (preferably include a link to your github page), and college-level math classes at the level of Calculus I or above.