Audio and Music Processing Lab (AMPLAB)

(Master in Sound and Music Computing, Universitat Pompeu Fabra)

General Information

Instructor: Enric Guaus

Schedule: Tuesday 18:00-21:00, Second term, 2011-12

Room: Ed. Tallers - Room: To announce

Course presentation

This course focuses on the state-of-the-art methods and tools for the automatic processing (description, generation and transformation) of music content.

Prerequisites

To take this course it is desirable to have an engineering background, to have taken some courses in Mathematics at the undergraduate level, like Algebra and Calculus, and also to be familiar with basic signal processing concepts. Programming experience and/or formal music knowledge is also desirable.

Competences to be acquired

General

Ability to find and use information and tools to solve a given practical problem
Ability to communicate in English
Ability to work in a team
Ability to work in an interdisciplinary field

Specific

Understand and use the current state of the art technologies for music processing
Apply music processing methodologies to solve practical problems

Grading

The course is organized in a set of lab sessions given by invited experts on different topics. First, the lecturer provides the required scientific background and, later, a practical problem is presented.
Sessions may last 1 or 2 weeks.
For each session, students are expected to review the lecture material and to read and write a short summary of some state-of-the-art papers related to the session.
There will be a set of proposed exercises or small programming assignments for each topic.
The grade is computed as the mean of all the assignments to deliver.
Last session is reserved to those students who want to present one of their assignments to all the students (an extra point is added to the final qualification).
All the assignments MUST be delivered.
A non presented assignment means you don't pass this course.

Schedule

#	Date	Topic	Lecturer	Abstract
0	10/01/2012	Presentation	Enric Guaus	In this session, we present the main goals, structure and grading conditions of the Audio and Music Processing Lab course.
1	10/01/2012	Python	Martín Haro and Dmitry Bogdanov	In this hands-on session we will introduce the basic characteristics of the Python programming language. Python is an interpreted programming language that allows you to do almost anything possible with a compiled language (C/C++/Fortran) without requiring all the complexity: Automatic garbage collection, Dynamic typing, Interpreted and interactive, Object-oriented, “Batteries Included”, Free, Portable, Easy to Learn and Use, Truly Modular.
2	17/01/2012	Essentia	Nicolas Wack	Essentia is an audio analysis tool that allows you to easily extract features from an audio signal and develop new ones suited to your taste. In this session we will examine how to work interactively in a python session with Essentia and extract a few musical descriptors.
3	24/01/2012	Gaia	Nicolas Wack	Gaia is a generic similarity search engine, which is mostly used for music similarity. In this session we will use some features extracted from audio files during the previous session about Essentia to implement a very simple genre classification algorithm.
4	31/01/2012	Canoris	Vincent Ackermans	In the first Canoris session we will look at the general theory of web applications and APIs and I will give a short overview of the service oriented architecture that powers Canoris. We will also discuss why these technologies are important to music technology. The session ends with some small assignments to learn how to interact with Canoris from Python.
5	07/02/2012	Canoris	Vincent Ackermans	In this second Canoris session we will move on to some bigger tasks and try to build a small application.
6	14/02/2012	FreeSound	Gerard Roma	Freesound.org is a massive repository of audio clips contributed by users under a CC license. The new web API makes it possible to interact with this database from a remote client. We will practice the basics of web-based audio hacking through an applied project.
7	21/02/2012	VST Plugins	Merlijn Blaauw	Learn basics of programming VST plug-ins in C++. We will go step-by-step through making 1) a gain effect, 2) a sine wave oscillator and 3) a resonant low-pass filter. Homework assignment is making a ping-pong delay effect. You can use classroom PC or own laptop (requires MSVC or XCode to be installed). Bring headphones.
8	28/02/2012	Weka: Explorer	Enric Guaus	Weka is a collection of machine learning algorithms for data mining tasks. Weka contains tools for data pre-processing, classification, regression, clustering, association rules, and visualization. It is also well-suited for developing new machine learning schemes. In this session, we focus on the use of Weka and how a general MIR problem should be addressed.
9	06/03/2012	Weka: Experimenter	Enric Guaus	In this session, we focus on the Experimenter application included in Weka to address multiple experiments and determine the relevance of different parameters of each one.
10	13/03/2012	Presentations	Enric Guaus	This last session is reserved to those students who want to present one of their assignments to all the students. Feedback from all the students is expected.

References

Each invited speaker will provide the specific list of references, webs and relevant software. All the documents will be available at Campus Global.

----
back to home