Study track
Every lecture course in deep-learning-drizzle (opens in a new tab), laid out so a topic can be picked, narrowed and watched without leaving the page. Choose a topic, then a course: it plays in the screen at the top of the page, which stays where it is while you keep looking.
Pick a topic below, then a course, and it plays here.
Loading the catalogue…
No course matches those filters.
Catalogue from kmario23/deep-learning-drizzle (opens in a new tab) (MIT), parsed by scripts/parse-drizzle.mjs into courses.json. Videos play from YouTube in privacy-enhanced mode; the courses and their recordings belong to their universities and instructors.
Questions
An index, not a course. Everything plays from the original publisher.
What is this?
Every lecture course listed in deep-learning-drizzle, laid out so a topic can be picked, narrowed by year or instructor, and watched without leaving the page. The player sits at the top and stays where it is, so choosing the next lecture never moves the thing being watched.
What is deep-learning-drizzle?
A public list, kept by kmario23 on GitHub, of university lecture courses on machine learning and the mathematics under it, with links to their recordings. It is MIT-licensed. scripts/parse-drizzle.mjs reads it and writes courses.json, which is what this page loads.
Who owns the videos?
The universities and instructors who recorded them. Nothing is rehosted here: the recordings play from YouTube in privacy-enhanced mode, and a course that is not embeddable opens on its own site instead. The filter for courses that play in the page exists because of that difference.
Why is it on an imaging site?
Because the imaging projects need it. The reconstruction, segmentation and registration work in this section rests on material from these courses, and keeping the track beside the projects makes it clear which parts were learned where rather than improvised.