I took the graduate course CS 268: Machine Perception with Professor Stefano Soatto, during Fall 2013. Overall, Professor Stefano does deeply care about teaching, and truly wants his students to learn the material. He has a unique knack of presenting material in the "right" way, i.e. in a good, intuitive manner. If you want to succeed in his course, you should be prepared to devote a *lot* of your time on his course. Like in many things: the more you put in, the more you get out. The course was extremely worthwhile, and I'm very glad I took it with Soatto. Due to the nature of the material, most of the lectures felt like a firehose of information. Coupled with the fact that Stefano tends to talk very quickly, I often found myself frantically scribbling notes during class, and then spending time outside of class reviewing my notes to solidify my understanding. Despite this, I often walked out of lecture feeling really excited about the course material - this can definitely be attributed to Soatto's lecture abilities. Fortunately, Stefano posts video lectures on the CourseWeb, which are the most valuable resources for this course. In these video lectures, Stefano re-explains the course material (along with his trusty tablet) in very good depth: most of these videos range from 45 - 80 minutes! When you're going over these, the pause button will be your best friend: be prepared to pause every 10 seconds, as you furrow your brow and try to understand what just happened, haha. Sure, watching a 45-minute video lecture turns into a ~1.5 hour affair - but you won't get left behind, and it really is a great learning resource. Even better is if you can watch the video lectures before the actual lecture. I will say that Stefano is more than a little intimidating at first. He is not afraid to be blunt or direct, which can be off-putting. However, do keep in mind that he does care about teaching, so don't be afraid to ask questions. Additionally, Stefano prefers to take a mathematically rigorous approach to things. Thus, if you wish to do well in his eyes, you must be prepared to brush up on your mathematics to really "speak" the correct language. But really - if you want to do Computer Vision, then this is a true necessity. As an aside - I loved Soatto's little stories. Every once in awhile, he'll tell funny (but interesting!) side stories about professors (such as how Professor Kalman was universally disliked), or the history of geometrically-correct perspective art in art history, etc.