Processamento da Fala

Why would you force students to go to the labs? Not everybody can be productive while beeing around 20 people screaming around in exercises where you are required to listen to audio files in order to make progress.

Course content is atractive, but the structure and the evaluation need to change. The course is basically 3 courses with 3 different professors. Labs are mandatory, which doesn't make any sense. Work load is high since each professor requires a report for its lab component. Exam is hard also, highest grade in my year was around 14. Even if you study hard, the way the exam is done, it will not convert into a great grade, because many are questions are mis-leading or focus on uncessary details of the course.

Contents

This course was effectively three mini-courses:

1. Signal processing and historical speech synthesis and modeling techniques, taught by Professor Luís Caldas de Oliveira;
2. Speech classification techniques, both historical and recent, taught by Professor Alberto Abad;
3. Modern speech processing systems and dialogue chatbots, taught by Professor Bruno Martins.

The signal processing section can be somewhat challenging...

I'm taking this course right now and have enjoyed it a lot. I was afraid to take it because few Computer Science students take it and I haven't taken LN yet, but it worked out fine.

Labs

The labs are manageable; There are 5 in total, 4 of them are in Python, easy and very interesting. Lab 2 was the toughest one, because it was very long and used MATLAB.

For some reason, out of the 5 labs, only 3 are randomly selected for evaluation.

Content

You learn a lot of things, namely phonetic alphabet, age classification, speech synthesis through audio concatenation and...