Animation and CGI Motion

I took this class as part of the Artificial Intelligence MicroMaster's program of edX. Why this class is part of an Artificial Intelligence curriculum is beyond me.

In a standard classroom setting I can see where this class is reasonable. When conducted within edX's online course system, however, the class becomes an enormous struggle. The code you must work with to complete your assignments is poorly documented and doesn't follow good coding practices. Feedback from TAs or anyone “in the know” is dodgy at best. You get one chance to submit your assignment yet have ambiguous information regarding how well your assignment will score prior to submission. Also, a submission must be done in multiple parts that has to be done in a particular order to get full credit. All in all, I feel as if the class was simply dumped into the edX online system with little regard to how well it translates from a classroom setting.

In my experience you'll need at least 20 hours a week if you're proficient with C++ programming, differential equations, and introductory college level physics. That number could double if you're rusty in any of these areas. Be wary if you take this class as part of the AI MicroMasters program as it could easily block you from getting the certificate after completing all the other coursework from the program. Take this course first, audit it, and if you can complete this course the other courses should be less time intensive and more related to artificial intelligence.

The course took much more time than the 8-10 hours per week advertised. Only a handful of us completed it, and the others had an experience similar to mine. It is unsuitable as a part-time MOOC that can be combined with full-time work. The course is technically and intellectually challenging, putting it at the right level for a postgraduate course in computer science. However, the basic problem is that scope of the work is too broad. No matter what one's skills as a computer scientist, there is too much project work to complete in a reasonable amount of time.

On the plus side, Professor Grinspun explains clearly and in an interesting way the concepts and methods that he is teaching. It combines a solid theoretical approach with thoroughly practical computational methods and one can imagine being able to put the whole thing to use in a practical production setting.