A collaboration between CINI Artificial Intelligence and Intelligence Systems (AI-IS) and The Mathworks for the development of free teaching materials on applied Artificial Intelligence
The goal of the project is to help students understand how diverse application domains may significantly benefit from the adoption of deep learning techniques by leveraging MATLAB as the foundation framework for the development of the relevant algorithms and techniques.
The project developed teaching material in the MATLAB environment that can be used for lessons and lab activities for all the courses facing digital signal processing in ICT and Engineering Faculties. The teaching material focuses on the use of Deep Learning techniques for multimedia signal processing, posing itself as a gentle introduction to the topic by mean of simple, but realist, use cases and applications. The material includes powerpoint slides, MATLAB Live Scripts, syllabus, and documentation to describe how to best reuse and adopt the content by other faculty/universities.
For each application domain, the project provides a set of modules, each having both theoretical and practical sections developed in the MATLAB environment, intended to gradually guide the student from basic concepts to advanced case-studies. All the examples are based on simple, but realistic, engineering problems and are addressed to both to bachelor’s and master’s students of ICT and Engineering Faculties. Following is a summary of each of the areas:
MULTIMEDIA SIGNAL PROCESSING (MMSP)
- Introduction to signal processing and Deep Learning What does signal processing means and implies, what deep learning is and how it can be used in the signal processing contest.
- Convolutional Neural Network for image analysis What CNN are and how they can be used for image processing and analysis
- Recurrent Deep Neural Network for video processing What RNN are and how expand image analysis techniques to video processing
- Audio signals processing Using Deep Neural Network for audio analysis and processing
The material is organized in folders, numbered according to the suggested reading order. Within each folder, the power point slides (if any) are intended to be read before the MATLAB livescript (if any). If the folder contains other files (such as .m, .mat, .csv, etc.) they are intended to support the execution of the livescript (thus there is no need to read/open them). If a livescript requires some toolboxes, they will be clearly pointed out in the introduction of the livescript.
For each slide there are some notes intended to support the teacher by giving they a possible lecture key. In the livescripts, suggestions intended only for the teacher are marked by [TEACHER NOTE: XX]. The livescripts need MATLAB >= R2020b.
A detailed list of lessons and topics addressed in this module is reported in the file Outline.xlsx, together with the expected teaching time. Please note that where the lesson name is to long for a file/folder, only a part of it was used (and the remaining part reported in brackets).