Sign in

End-to-End Learning of Visual Representations from Uncurated Instructional Videos

By Antoine Miech and others
Annotating videos is cumbersome, expensive and not scalable. Yet, many strong video models still rely on manually annotated data. With the recent introduction of the HowTo100M dataset, narrated videos now offer the possibility of learning video representations without manual supervision. In this work we propose a new learning approach, MIL-NCE,... Show more
January 17, 2020
=
0
Loading PDF…
Loading full text...
Similar articles
Loading recommendations...
=
0
x1
End-to-End Learning of Visual Representations from Uncurated Instructional Videos
Click on play to start listening