Program Overview

Master in Sound and Music Computing

The Master in Sound and Music Computing is a program that focuses on the application of machine learning and other techniques to sound and music analysis.

Overview

The program provides an overview of common machine learning techniques used in sound and music audio analysis.

Academic Program

The academic program includes a course on Machine Learning for Sound and Music, which is offered in 10 weeks with 25 hours of lectures and hands-on sessions. The course covers various topics, including:

• Introduction to ML and AI
• Linear/logistic regression and gradient descent
• Traditional machine learning algorithms (kNN, decision trees, etc.)
• Artificial neural networks
• CNNs, audio representations
• Audio autoencoders
• Metric learning for audio
• Audio transformers
• Audio embedding models and transfer learning applications
• Text-audio models

The course is taught by instructors Dmitry Bogdanov, Rafael Ramirez, and Pablo Alonso, and is worth 5 ECTS credits. The evaluation of students is based on weekly assignments and participation in class.

Materials and References

The course materials include slides and relevant code, which are made available before each class. The practical work is conducted using Python, including libraries such as scikit-learn, Keras, PyTorch, PyTorch Lightning, and TorchAudio. The course references include:

• Aurélien Géron, (2020). Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow, 2nd Edition, O’Reilly Media.