Trackverse: a Large-Scale Object-Centric Video Dataset for Image-Level Representation Learning.

Yibing Wei, Samuel Church, Victor Suciu, Jinhong Lin, Cheng-En Wu

Browse the full ICCV paper archive.