Overall this is a pretty chill class but you also learn a lot. You don't have to have ML experience coming into the class, and don't have to start working on the final project until Scalzo covers ML in class, which is nice. Workload is a couple homework assignments done in Google Colab using Python libraries to do image processing and stuff, plus the group project. Some of the homework problems seem tricky but the TAs do demos with extremely similar problems in discussion so it ends up being pretty straightforward.

Most of the grade is the group project, which you work on for the later half of the quarter. You have a lot of leeway in choosing your topic and can have an advisor to help (mine was Scalzo and he was really good about explaining each part of the project and giving feedback on our progress). Scalzo doesn't grade it too strictly, but you do have to put effort into it.