Mathematics

Schedule an appointment

Video Art with FFMPEG is a course for artists interested in unlocking the power of ffmpeg to create generative video projects. In this course we will use Google Colab to combine Python and ffmpeg to create scriptable video tools. We’ll also look at ffmpeg-python, a modern coder-friendly syntax that can be used as an alternate scripting language for ffmpeg. If you’ve ever used an ffmpeg script and found it powerful but utterly confusing, I think you’ll enjoy this class.Note: This is not a machine learning class, although you are welcome to use video you generate with ML for this course.PrerequisitesYou should be comfortable with the basics of coding. Experience with Python is helpful but not requiredAccess to a Windows, Mac, or Linux machine that can run ZoomA Google account that can use Google Colab (if you are uninterested in using Google tools, notebooks can be downloaded and used in Jupyter notebooks or modified to work in a traditional Python environment)Example syllabus (subject to change)Week 1: ffmpeg basicsWeek 2: complex filter operationsWeek 3: ffmpeg-python, scriptable transformationsWeek 4: free space/project examplesSchedule and StructureFull CourseRunwayML + Google Colab will meet live for 4 weeks on Saturdays at noon EST starting on May 22nd. You should be able to attend every session via Zoom, so please make sure your schedule allows for it. In addition to the four 2-hour sessions, there will be two open 1hr sessions with Derrick every week to answer questions and get help on projects. We will be using Slack for asynchronous messaging and discussion. Additional materials outside of the lectures and demos will be provided as video links, webpages or PDFs.Artificial Images classes offer you individual time with both instructors and small class sizes (~20 students). These classes have been taught to students across the world and we’re constantly iterating to improve them and cover the latest techniques in machine learning.Audit OptionFor those who cannot attend in-person or prefer to learn solely from recordings, we’re now offering an audit option. You will be sent links to each class recording as soon as available (usually within 24 hours) and will have a private Slack channel where you can ask questions. About the instructorsDerrick Schultz is a designer and artist working with machine learning. His work has been featured at machine learning conference CVPR and NeurIPS, and has been commissioned by the New York Times, HP, and Acne. He has taught numerous machine learning courses, and has teaching experience at ITP, Parsons, and SVA. You can see more of his work on his website.