๐ŸŽ“How I Study AIHISA
๐Ÿ“–Read
๐Ÿ“„Papers๐Ÿ“ฐBlogs๐ŸŽฌCourses
๐Ÿ’กLearn
๐Ÿ›ค๏ธPaths๐Ÿ“šTopics๐Ÿ’กConcepts๐ŸŽดShorts
๐ŸŽฏPractice
๐ŸงฉProblems๐ŸŽฏPrompts๐Ÿง Review
Search
How I Study AI - Learn AI Papers & Lectures the Easy Way

Papers1

AllBeginnerIntermediateAdvanced
All SourcesarXiv
#data attribution

Motion Attribution for Video Generation

Intermediate
Xindi Wu, Despoina Paschalidou et al.Jan 13arXiv

Motive is a new way to figure out which training videos teach an AI how to move things realistically, not just how they look.

#motion attribution#video diffusion#optical flow